Retour

Explorez tous les épisodes du podcast Impact AI

Plongez dans la liste complète des épisodes de Impact AI. Chaque épisode est catalogué accompagné de descriptions détaillées, ce qui facilite la recherche et l'exploration de sujets spécifiques. Suivez tous les épisodes de votre podcast préféré et ne manquez aucun contenu pertinent.

Rows per page:

1–50 of 127

TitreDateDurée
Foundation Model Series: Transforming Biology with Zelda Mariet from Bioptimus17 Feb 202500:21:26

Zelda Mariet, Co-Founder and Principal Research Scientist at Bioptimus, joins me to continue our series of conversations on the vast possibilities and diverse applications of foundation models. Today’s discussion focuses on how foundation models are transforming biology. Zelda shares insights into Bioptimus’ work and why it’s so critical in this field. She breaks down the three core components involved in building these models and explains what sets their histopathology model apart from the many others being published today. They also explore the methodology for properly benchmarking the quality and performance of foundation models, Bioptimus’ strategy for commercializing its technology, and much more. To learn more about Bioptimus, their plans beyond pathology, and the impact they hope to make in the next three to five years, tune in now.


Key Points:

  • Who is Zelda Mariet and what led her to create Bioptimus. 
  • What Bioptimus does and why it’s so important.
  • Why their first model announced was for pathology.
  • Zelda breaks down three core components that go into building a foundation model.
  • How their histopathology foundation model is different from the number of other models published at this point.
  • Their methodology behind properly benchmarking how well their foundation model performs.
  • Different challenges they’ve encountered on their foundation model journey.
  • How they plan to commercialize their technology at Bioptimus. 
  • Thoughts on whether open source is part of their long-term strategy for the model, and why.  
  • Developing a product roadmap for a foundation model.
  • She shares some information regarding their next step, beyond pathology, at Bioptimus.
  • The importance of understanding what kind of structure you want to capture in your data.
  • Where she sees the impact of Bioptimus in the next three to five years. 


Quotes:

“Working on biological data became a little bit of a fascination of mine because I was so instinctively annoyed at how hard it was to do.” — Zelda Mariet


Bioptimus is building foundation models for biology. Foundation models are essentially machine learning models that take an extremely long time to train [and] are trained over an incredible amount of data.” — Zelda Mariet


“There are two things that are well-known about foundation models, they’re hungry in terms of data and they’re hungry in terms of compute.” — Zelda Mariet


“On the philosophical side, science is something that progresses as a community, and as much as we have, what I would say is a frankly amazing team at Bioptimus, we don’t have a monopoly on people who understand the problems we’re trying to solve. And having our model be accessible is one way to gain access into the broader community to get insight and to help people who want to use our models, get insight into maybe where we’re not doing as well that we need to improve.” — Zelda Mariet


Links:

Zelda Mariet on LinkedIn

Zelda Mariet

Bioptimus


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Foundation Model Series: Democratizing Time Series Data Analysis with Max Mergenthaler Canseco from Nixtla10 Feb 202500:27:11

What if the hidden patterns of time series data could be unlocked to predict the future with remarkable accuracy? In this episode of Impact AI, I sit down with Max Mergenthaler Canseco to discuss democratizing time series data analysis through the development of foundation models. Max is the CEO and co-founder of Nixtla, a company specializing in time series research and deployment, aiming to democratize access to advanced predictive insights across various industries.

In our conversation, we explore the significance of time series data in real-world applications, the evolution of time series forecasting, and the shift away from traditional econometric models to the development of TimeGPT. Learn about the challenges faced in building foundation models for time series and a time series model’s practical applications across industries. Discover the future of time series models, the integration of multimodal data, scaling challenges, and the potential for greater adoption in both small businesses and large enterprises. Max also shares Nixtla’s vision for becoming the go-to solution for time series analysis and offers advice to leaders of AI-powered startups.


Key Points:

  • Max's background in philosophy, his transition to machine learning, and his path to Nixtla.
  • Why time series data is the “DNA of the world” and its role in businesses and institutions.
  • Nixtla's advanced forecasting algorithms, the benefits, and their application to industry.
  • Historical overview of time series forecasting and the development of modern approaches.
  • Learn about the advantages of foundation models for scalability, speed, and ease of use.
  • Uncover the range of datasets used to train Nixtla's foundation models and their sources.
  • Similarities and differences between training TimeGPT and large language models (LLMs).
  • Hear about the main challenges of building time series foundation models for forecasting. 
  • How Nixtla ensures the quality of its models and the limitations of conventional benchmarks.
  • Explore the gap between benchmark performance and effectiveness in the real world.
  • He shares the current and upcoming plans for Nixtla and its TimeGPT foundation model. 
  • He shares his predictions for the future of time series foundation models.
  • Advice for leaders of AI-powered startups and what impact he aims to make with Nixtla.


Quotes:

“Time series are in one aspect, the DNA of the world.” — Max Mergenthaler Canseco


“Time is an essential component to understand a change of course, but also to understand our reality. So, time series is maybe a somewhat technical term for a very familiar aspect of our reality.” — Max Mergenthaler Canseco


“Given that we are all training on massive amounts of data and some of us are not disclosing which datasets we’re using, it’s always a problem for academics to try to benchmark foundation models because there might be leakage.” — Max Mergenthaler Canseco


“That’s an interesting aspect of foundation models in time series, that benchmarking is not as straightforward as one might think.” — Max Mergenthaler Canseco


“I think right now in our field probably benchmarks are not necessarily indicative of how well a model is going to perform in real-world data.” — Max Mergenthaler Canseco


“I think that we’re also going to see some of those intuitions that come from the LLM field translated into the time series field soon.” — Max Mergenthaler Canseco


Links:

Max Mergenthaler Canseco on LinkedIn

Nixtla

Nixtla on X

Nixtla on LinkedIn

Nixtla on GitHub


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Decoding the Immune System for Drug Discovery with Noam Solomon from Immunai04 Nov 202400:18:02

Today’s guest believes that decoding the immune system is at the heart of improving drug efficacy. He is currently focused on this effort as the CEO and Co-founder of Immunai – a company that is building an AI model of the immune system to facilitate the development of next-generation immunomodulatory therapeutics. Noam Solomon begins our conversation by detailing his professional history and how it led to Immunai before explaining what Immunai does and why this work is vital for healthcare. Then, we discover how understanding the immune system will help to improve how drugs work in our bodies, how the team at Immunai accomplishes its goals, the major challenges of working with complex ML models, and some helpful recommendations for processing the high-dimensional nature of biological data. Noam also explains the collaborative landscape of Immunai, how the evolution of technology made his work possible, Immunai’s plans for the future, and his advice to others on a similar career path. 


Key Points:

  • Unpacking Noam Solomon’s professional journey that led to his founding of Immunai. 
  • What Immunai does and why this work is vital for the healthcare industry. 
  • How understanding the immune system will help to improve drug efficacy. 
  • Exploring how Noam and his team use AI to accomplish their goals. 
  • The standardization of data and other challenges of working with complex ML models. 
  • Techniques for handling the high-dimensional nature of biological data.
  • How ML experts collaborate with other domains to inform and build Immunai’s models. 
  • The technical advancements that have made Noam’s work possible. 
  • His advice to other leaders of AI-powered startups, and imagining the future of Immunai. 
  • How to connect with Noam and his work.  


Quotes:

“First, let’s talk about the problem, which is today, getting a drug from IND approval to FDA approval—which is the process of doing clinical trials—has less than a 10% chance of success, usually about a 5% chance, takes more than 10 years, and more than $2 billion of open immune therapy.” — Noam Solomon


“Different people respond differently to the same drug, and the reason they respond differently is because their immune system is different.” — Noam Solomon


“You first need to fall in love with the problems. Many ML people—physicists, mathematicians, computer scientists—we love building models; we love solving puzzles. In biology, you need to really fall in love with the question you are trying to answer.” — Noam Solomon


“It’s a great decade for biology.” — Noam Solomon


Links:

Noam Solomon on LinkedIn

Noam Solomon on X

Immunai


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Detecting Breast Cancer Earlier with Tobias Rijken from Kheiron Medical30 Jan 202300:27:55

All women face the risk of breast cancer, but early detection can greatly increase the chances of a positive outcome and reduce the need for aggressive treatment options. In this episode, I talk with Tobias Rijken, CTO and co-founder of Kheiron Medical Technologies, about leveraging AI for detecting breast cancer. We discuss the role of AI in improving medical care, the power of vertical integration and feedback loops, and what makes Kheiron different from other AI startups. Hear about the challenges of acquiring reliable data, whether using generative models is beneficial, details about the products Kheiron has created, and much more!


Key Points:

  • Tobias's professional background and why he created Kheiron Medical Technologies.
  • Learn about the amazing work Kheiron Medical Technologies does and why it is important.
  • Overview of why detecting breast cancer early is so vital and the challenges of screening.
  • How AI can help resolve the current challenges in cancer screening.
  • He explains the machine learning process and training the model used.
  • The complications encountered in working with radiology images.
  • Find out why image quality is key to the machine learning process.
  • How he is able to account for the variation of technology and methods used.
  • Outline of the regulatory process and how it impacts machine learning model development.
  • Hear advice Tobias has for other leaders of AI-powered startups.
  • Details about how Tobias approaches improving the models over time.
  • Tobias tells us what Kheiron Medical Technologies has planned for the future.


Quotes:

“What I liked so much about machine learning is the ability it has to solve real-world problems. And in my opinion, real-world machine learning is very different from academic machine learning.”


“Either the right information isn't available, or it is inaccurate, or there's missing information. We see AI as a tool to help address those information problems.”


“The challenge when you sample uniformly from your whole dataset is that there will be cases you've sampled, where you may not have ground truth.”


“For me, when I started this company, this was not about building a great model that has a great performance on a test dataset. This is about getting AI into the real world.”


Links:

Tobias Rijken on LinkedIn

Tobias Rijken on Twitter

Kheiron Medical Technologies

Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Computer Vision Advisory Services – Monthly advisory services to help you strategically plan your CV/ML capabilities, reduce the trial-and-error of model development, and get to market faster.

Monitoring Fields for Precision Agriculture with Gershom Kutliroff from Taranis23 Jan 202300:34:39

In this episode, I talk with Gershom Kutliroff, CTO of Taranis, about precision agriculture. Taranis uses computer vision to monitor fields, providing critical insights to growers. Gershom and I talked about how they gather and annotate data, the challenges they encounter in working with aerial imagery, and how they validate their models and accommodate data drift with continuous learning.


Quotes:

“Taranis is using drone technology to capture imagery and then use AI to process that imagery to understand what's happening in grower's field.”


“It becomes increasingly difficult to maintain consistent quality levels if you're working with tens or even hundreds of annotators. But when you have AI models, then you have the ability to control the quality of the insights that you're generating.”


“There's a lot of discussion in the last few years in the AI space about data-centric versus model-centric. Model centric would be the case where in your development you invest a lot in  choosing the right architecture that optimizes your performance, gives you the best results for your models, or spending a lot of time with hyper parameters and that type of work. And data-centric is you spend a lot more time making sure that your data set is clean, that you've got that it's balanced, you've got the right amount of classes for the problem that you're trying to solve.”


“We struggle with the problem of long tail distributions. If I take diseases as an example, there are some diseases that can cause a lot of damage to the crops. But they're very rare in terms of how often they actually occur in grower's fields.”


“Because we're running our own operations and so we're flying our own drones, we've also  invested in the software that's running on the drones when we're flying. So the images the drone pilot captures in the field are validated in the field. We have algorithms running on the edge to be able to check the quality of those images. And then if the images are not the quality that we expect them to get, the pilot knows while he's still there at the field and he can fly again.”


“For a lot of the models that we use you really need domain experts. You really need trained agronomists who can look at these images.”


“A certain percentage of all of the missions that we've flown are sent for review by our in-house agronomists before we release them to customers. So that's a really critical piece of how we do validation, and that also gives us a high level of confidence internally that the product that we're releasing to our customers stands by the quality that we expect it to.”


“We do suffer from this type of data drift where the data that we're seeing in production is not exactly in the same distribution as the data that we used to train. So the most effective technique that we've seen is to implement some kind of a continuous learning type of framework whereby we are able to take data that we're capturing in production, so when we're actually live   and servicing our customers' fields. And then the data that doesn't have a good correspondence with the distribution of the training data that was used for the models, we can then filter that data out. We can extract that data and use it to quickly retrain the models, to adapt the models, and then deploy those models back into production.”


“The company started by offering a product based on manual tagging, which didn't have any AI technology at the beginning, which allows it to offer products and service customers and start building this very rich database that we leverage now.”


Links:

Taranis

Gerhsom Kutliroff

Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Computer Vision Advisory Services – Monthly advisory services to help you strategically plan your CV/ML capabilities, reduce the trial-and-error of model development, and get to market faster.

Improving Patient Outcomes with Vinod Subramanian from Syapse16 Jan 202300:35:52

In this episode, I talk with Vinod Subramanian, Chief Data and Product Development Officer at Syapse, about machine learning for healthcare and advancements in cancer treatment. Syapse is a real world evidence company dedicated to improving outcomes for cancer and other serious diseases. Vinod and I talked about the types of healthcare data they work with, the data challenges they encounter, how they validate their models, and how they mitigate bias.

QUOTES:
"Technology is not the answer exemplified of the intent. And the fundamental question, I think, that all of us are confronted by: what is the intent and what in the world that we want to try to help shape?"

"There are infinite possibilities in the terms of patient care with aggregated and harmonized data in healthcare. We all know about the point that data in general is fragmented and decentralized in the industry. Real world data comes from knowledge and knowledge comes from collecting information and of course, information stems from aggregating disparate data."

"Machine learning today, especially in a life science setting, is leveraged as new ways right to garner new biological insights."

"One of the things that we are also doing is not just about adopting and using (ML and NLP), we strongly believe that we want to share our work. And that would not only raise and mainstream the work of everybody doing it, but also it'll help us in adopting and applying in precision medicine through standards."

"Now not all data is needed equal. When we can improve the way data is collected, connected, analyzed, and consumed, we can not only improve the lives of our community, but it also gives us a way to look at the care continuum very differently."

"There's no guarantee when you get into an initiative which uses machine learning and AI, because it cannot be successful. It has to be a learning experience, but it, there's no guarantee that it will be  successful. And there needs to be willingness and appetite to experiment, learn, and iterate, and taking a Socratic approach, and accelerate the journey towards success, anchor down the culture."

LINKS:
Syapse
Vinod Subramanian

Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Computer Vision Advisory Services – Monthly advisory services to help you strategically plan your CV/ML capabilities, reduce the trial-and-error of model development, and get to market faster.

Ecological Restoration with Patrick Leung from Earthshot Labs09 Jan 202300:28:08

In this episode, I talk with Patrick Leung, Co-founder and CTO of Earthshot Labs, about using machine learning to help predict and restore forests and our ecosystem. Earthshot Labs is building the technology and expert guidance to develop and finance nature-based carbon projects globally. Patrick and I talked about how Earthshot Labs gathers and annotates data, the challenges in working with remote sensing and other forms of data, the importance of collaboration across disciplines, and how machine learning tools can help save our ecosystems.

QUOTES:
"We are able to actually bridge that financing gap and unlock a whole bunch of new projects that can then be in the carbon marketplace, and also bring a host of benefits to both the ecosystem, as well as, the communities that live around the ecosystem."

"Machine learning is really essential because what we're trying to do here is predict the future. We're trying to predict the next 30 years of a forest regrowing in a tropical region."

"We must look at the past. We must look at whatever data we can gather from the past state of the ecosystem and use various machine learning methods to predict the future in order to provide a view on what's gonna happen on this land in the future when we do this project."

"These are actual mathematical simulations that take into account the current conditions of the ecosystem and actually forecast them by using a kind of simulation that incorporates photosynthesis and evapotranspiration and other forms of ecological processes."

"They would look at historical flood maps and essentially combine them with flood forecasting models i order to generate what is a given area going to look like if it gets flooded in the future because of climate change or for other reasons. And I was very enamored with that. I thought that was a very, very clever use of a technology."

"I think what we're doing definitely encompasses biodiverse native ecosystems and just restoring as many of them as we can throughout the most critical parts of the biosphere, that there are in this world. And also helping to switch our societal systems into more of a harmonious, and regenerative relationship with those ecosystems."

LINKS:
Earthshot Labs
Patrick Leung

Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Computer Vision Advisory Services – Monthly advisory services to help you strategically plan your CV/ML capabilities, reduce the trial-and-error of model development, and get to market faster.

Personalized Physiology Analytics with Matt Pipke from physIQ02 Jan 202300:33:30

In this episode, I talk with Matt Pipke, Co-founder and Chief Digital Health Officer of physIQ, about personalized medical predictions from physiology data. Matt and I talked about the challenges in working with physiology data, how to validate models and minimize bias, and the importance of collaboration.


Quotes:

“What physIQ does is it harvests data from those continuous data streams from wearable sensors and produces analytical results that are useful for clinical care when taking care of patients who are outside the four walls of the hospital and in scientific endeavors such as clinical trials where it's interesting to know what the efficacy of the drug is on a target disease, whether the health of the patients who might take those drugs is being improved or at least is not degrading any further.”


“What we have to do is build our algorithms and our analytics based on machine learning techniques and, of course, the more recent really successful subgroup of deep neural net algorithms that can sift through this data and can highlight accurately the vital signs of physiology we need to make the assessments available.”


“So part of the issue there, is to figure out how to differentiate the background variation that's normal for people as they move around in their daily lives from the telltale signs that they may be suffering from a derangement of physiology.”


“There's a lot of companies and offerings out there that are in the consumer fitness market. They might be appropriate for healthy populations that are looking to track their activity, the amount of sleep that a healthy person might get, but they're really not the right target populations of interest for the medical system or for clinical trials where you have a population that's suffering from a disease that a drug is targeting.”


“Now I know that a lot of companies out there tend to avoid the regulatory pathway for medical or health or fitness applications, and I don't think that's a good move. . . The FDA experience for us has been at times frustrating of course, as it is for anybody who has to deal with regulations, but at the same time, there is a core of meaningful value add there. Regulatory agencies around the world, FDA included, they have a pretty thankless job. They never get credit for what they do. They only get complained about. But what they're doing is really, really critical to outputting valuable, usable product in the healthcare and medical space.”


“So bias in models really comes back to the representativeness of your data, right? So if you've got data that's not representing the target users, the target populations that you're going to analyze, you can end up with bias. You can end up with bias in surprising ways.”


“If you aren't aware of what might be lurking in your data, you could be overfitting the wrong thing and then find out that your algorithm does not generalize, does not work in other areas.”


“My feeling about this is that it's all about the data. physIQ got started a lot earlier than we probably should have and we've benefited in a strange way in that we've been in the game a lot longer than other players in this space. So we've been collecting data for a long time and we built a robust platform to collect data.”


“There's a lot of resistance to change and, in fact, the layperson might be horrified to learn how the healthcare system actually works. But, stepping back, something definitely has to change in healthcare. We all know that it's not sustainable the way things are now. But we don't have any illusions at physIQ about how a little company like ours can change things by ourselves. It's really about timing. Right. And sometimes you have to look for those windows of opportunity when in large industries with huge amounts of existing business relationships and the way that they work today are ready for change.”


Links:

physIQ

Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Computer Vision Advisory Services – Monthly advisory services to help you strategically plan your CV/ML capabilities, reduce the trial-and-error of model development, and get to market faster.

Early Cancer Detection with Emi Gal from Ezra19 Dec 202200:23:57

In this episode, I talk with Emi Gal, co-founder and CEO of Ezra, about cancer screening with a full body MRI scan. Ezra is on a mission to detect cancer early for everyone by making the process more accurate, faster, and cheaper. Emi and I talked about the challenges in working with MR data, how regulatory processes affect model development, and the importance of validation.


Quotes:

“What we've been able to achieve is to essentially reduce the cost and the time in a scanner of an MRI from about two to three hours for full body to 60 minutes. And we're actually working on a new AI that will roll out next year that will reduce the scan time to 30 minutes.”


“What we do is we acquire the scan fewer times, and then we've built machine learning models that recognize what noise looks like and then just remove that noise. And then we kind of expanded that from not just noise. If you acquire scans with lower resolution, the resulting images are a little bit blurry so we can sharpen them.”


“Our focus on the scanning front is to reduce scan time, which yields these images with increased noise artifacts, and then use machine learning to enhance these images so that a radiologist can then use them for interpretation.”


“I think what having to receive FDA clearance for AI does, is it really forces the company from day one to think about what are all of the things that might influence the performance of said AI, and what can we do to ensure that we maximize the chances of success?”


“We have had an instance when we had to go back to the drawing board and build the model again because we failed internal validation prior to formal validation that we had to submit to the FDA.”


“I think the way you ensure that the technology we develop fits the clinical workflow is actually not starting with the technology, but starting with the end goal in mind and then figuring out what you need to do in order to achieve that.”


“To screen a hundred million people a year, we think, is a huge endeavor and probably going to take a decade or two to achieve. And I'm personally committed to Ezra for the rest of my career.  In the next three to five years, I would hope we are making good progress towards that mission, and maybe in five years we're screening at least a million people a year.”


Links:

Ezra

Emi Gal

Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Computer Vision Advisory Services – Monthly advisory services to help you strategically plan your CV/ML capabilities, reduce the trial-and-error of model development, and get to market faster.

Sorting Recyclables with Amanda Marrs from AMP Robotics12 Dec 202200:21:15

In this episode, I talk with Amanda Marrs, senior director of product at AMP Robotics about modernizing the world’s recycling infrastructure. Amanda and I talked about how they ensure their models work for a diverse set of objects, measuring the success of their technology, and some tips for building a successful ML team.


Quotes:

“At AMP we have a broad mission of enabling a world without waste.”


“We work backwards on everything that ends up in a landfill to develop the technology we need to keep that from happening.”


“We really have two main areas that we work in. One is technology that we will put in place at a material recovery facility. . . The other half of what we do at AMP is use our own technology for what's called a secondary sortation facility.”


“All of this technology really has three main components. You have to be able to see the objects on the belt, and that's where the machine learning comes in. You have to be able to sort the objects effectively, and there's some ML behind that as well. And then you have to be able to report, see what's happening, and draw conclusions and make decisions and optimize further in the facilities.”


“A majority of the data fits nicely within these primary categories. But, in AI, typically there's this natural long tail, and we have that as well.”


“Diversity is the name of the game in this industry where you have to be able to recognize everything. And so a huge sample set of data really helps us overcome that.”


“The wonderful thing about AI, it doesn't get tired, it doesn't get dizzy. And it can keep its inference at the same rate.”


“What we try to do when we translate this to customers, to non deeply technical folks – they're technical in other ways, but they're not dealing with AI all day – is we really try to translate it to the outcomes.”


“Start your hiring process early so that you're expecting it might take a while before you really, really need that team member joined, onboarded, trained up and enabled to help deliver on projects.”


“I think, for us, recruiting and thinking about what mix of talent we really need on a team, it's looking across all of those different areas and building out a team that really compliments each other's skillsets.”


Links:

AMP Robotics

Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Computer Vision Advisory Services – Monthly advisory services to help you strategically plan your CV/ML capabilities, reduce the trial-and-error of model development, and get to market faster.

Cell Sorting with Mahyar Salek from Deepcell05 Dec 202200:32:04

In this episode, I talk with Mahyar Salek, co-founder and CTO of Deepcell, about an AI powered technology for single cell analysis through the lense of high content cell morphology.  Deepcell's platform blends deep learning, microfluidics, and high resolution optics to deliver novel insights about cell biology and has the capability to sort, label-free for downstream multi-omic and functional analysis for use in research, translational studies, and therapeutic research.  We discussed some of the challenges and opportunities in working with single cell images and how they used self-supervised learning.


Quotes:


“We really use the power of computer vision and AI capabilities combined with the advances in microfluidics and imaging to create this high dimensional, high content interpretation of single cell images. And we use that in real time to purify and separate cells of interest.”


“We have to see millions of cells even in just one go, one run. So you can't really do that without the scalability of an algorithm, right? And then we have to be consistent and robust.”


“When I hear challenges, I equate them with opportunities and I'll tell you why. So, for instance, one of the challenges, not just with us, but any sort of AI solution that looks at biological samples is the susceptibility to artifacts.”


“But as soon as you roll it out, there's a difference between your lab and the lab, you know, a block down the road because of the artifacts. So it's artifacts are definitely challenging, but for us, it's an opportunity as I mentioned, because we generate the data through our own platform and that means that we have a very controlled environment.”


“Because, again, we have the full control over the imaging path and where the cells lie, where we image them, we could actually do these sort of things and come up with models that are very less reliant on labels.”


“By being able to run a biological assay and validate whether the existing model, like basically errors in the existing models and existing labels, and that way you're able to iterate very quickly on your learning without even relying on arguably erroneous human labels, erroneous and obviously expensive human labels.”


“Any modern life science companies that rely on data, you have to have a very tight collaboration between machine learning and data scientists and the domain experts.”


“It is really important to, as you kind of come up with a development strategy and the product strategy, understand where you could rely on AI today versus where you hope that the AI could deliver, you know, two years down the road.”


Links:

Deepcell

Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Computer Vision Advisory Services – Monthly advisory services to help you strategically plan your CV/ML capabilities, reduce the trial-and-error of model development, and get to market faster.

Data-driven Pathology with Coleman Stavish and Julianna Ianni from Proscia28 Nov 202200:30:32

In this episode, I talk with Coleman Stavish and Julianna Ianni from Proscia about data-driven pathology. Coleman is the co-founder and CTO of Proscia and Julianna is the VP of AI Research & Development. We discussed the importance of quality control systems in an ML pipeline, model generalizability, and how the regulatory process affects ML development.


Quotes:


“Better accuracy in diagnosis means less overdiagnosis and less under diagnosis, which typically leads to better patient outcomes and quality of life.”


“Pathology is crucial in the drug development pipeline. It's helping pharmaceutical companies develop new treatments while assessing their safety and efficacy.”


“You'll often find slides that have been annotated with pen ink. That's something that can be quite common to do in some settings and that, if you're trying to train a diagnostic model, can really bias the model.”


“One of the heaviest impacts to development for us, just to give you an example, has been areas where we find a great level of disagreement in the ground truth data. So that will come out when you test, and we have to account for that disagreement during development.”


“It also requires thinking through, not just how are we going to validate, but then how are we going to keep tabs on the different deployments and ensure that we're not seeing performance degrade as maybe the data or the conditions within the laboratory change.”


“No matter how accurate or how valuable that information is that's produced by the model, if it's not actually introduced in the right way into the overall workflow, it's not going to be put into routine use.”


“Prepare to iterate. A solution that you build is probably not going to be the final destination, the final solution. And I think the fast pace of this field kind of demands some constant innovation.”


“I'd also say to heavily invest in your team. There's really nothing that replaces having good people and very skilled people working for you and building these AI products.”


“Something that we've learned ourselves is how to balance the investor pitch about AI and its potential with the near and immediate term. Smaller successes that build you a road to that more ambitious future.”


“They could have the ability to diagnose cases remotely without having and maybe assisting patients who are in far flung areas of the world that may not have access to subspecialty pathologist expertise.”


“Maybe it means someone gets the right diagnosis a little bit faster in aggregate. I think that could have a really big impact.”


Links:

Proscia

Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Computer Vision Advisory Services – Monthly advisory services to help you strategically plan your CV/ML capabilities, reduce the trial-and-error of model development, and get to market faster.

Biophysical Modeling of Cancer with Joe Peterson from SimBioSys14 Nov 202200:50:13

In this episode, I talk with Joe Peterson, co-founder and CTO of SimBioSys, about biophysical modeling of cancer. SimBioSys is trying to revolutionize precision cancer care through individualized treatment planning, accelerated drug development, clinical trial optimization, and comprehensive biomarker development. Joe and I talked about the challenges of working with heterogeneous forms of data and the ways bias can manifest when training models on medical data.


Quotes:

“We use AI or ML at effectively every point in the process, both in our clinical medical devices, but also for our internal R&D.”


“Have you ever seen the way weather scientists simulate a hurricane? We do a very similar thing within the body, or if you've ever seen mechanical engineers simulate the combustion of a gas and a gas turbine, we do a similar type of thing within these patient models.”


“If you're able to distill the processes that go on biologically, chemically and physically to their essence, you can create building blocks that can be mixed and matched.”


“Our thought was, let's not ask the models to do too much. Let's ask them to do one thing that we need them to do very, very well. This allows us to have more collected data or more directed data collection, as well as more clearly defined goals in terms of business value and delivering business value to each of the models.”


“All these different types of data are much more heterogeneous. They come from many different scales. They come from many different sources. They're encoded in many different ways, and so there's a huge effort, on the research and development side, just to extract what's meaningful in those different types of data sets so that we can begin to define those biophysical building blocks that ultimately make it into the clinical application.”


“It's just really about capturing the variability and trying to drive out as much variability up front as you possibly can.”


“We also develop models that are generally capturing any sort of drift in the data over time.”


“You wanna understand outside of just a research setting, but out there in the wild how well your models are going to work, how often you're going to return a null result or an inconclusive result to a physician and being able to track that over time is really important from a quality control standpoint.”


“It's all the quality control machine learning models and deep learning models that make up the bulk of those internally.”


“Our responsibility as practitioners of AI is to not only identify and understand that bias, that historical bias, but also try to account for it as best we can.”


“What we need to assess when developing drugs or algorithms or devices is how they were trained, how they were tested, and really stratify those patient populations as best we can to sort of understand, at the very least, how they're behaving.”


“We've spent a lot of time trying to account for that variability as best we can. That said, we don't have a perfect data set and we're constantly thinking about ways to improve it.”


“I think what it comes down to is being open and transparent and really looking at the data that you have at the end of the day, If doctors are going to trust medical devices and if they're going to trust AI, they need to have information about.”


“By looking into and stratifying the patient populations in that way we can better understand where we need to targetedly spend resources to collect potentially more data to better understand the performance in those places or to improve our algorithms.”


“Adopt good machine learning practices early, just like good clinical practice or good manufacturing practices that are standards that are now being drafted and adopted.”


“Find the right partners to sort of drive the questions that you're addressing and ultimately the clinical actions that you're trying to address.”


“Models that are built to do a single task excellently well is a better approach than trying to build a model that does four or five tasks really well.”


Links:

SimBioSys

Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Computer Vision Advisory Services – Monthly advisory services to help you strategically plan your CV/ML capabilities, reduce the trial-and-error of model development, and get to market faster.

Foundation Model Series: Accelerating Radiology with Robert Bakos from HOPPR28 Oct 202400:28:42

Imagine a world where radiology backlogs are a thing of the past, and AI seamlessly augments the expertise of radiologists. Today, I'm joined by Robert Bakos, Co-Founder and CTO of HOPPR, to discuss how his company is bringing this vision to life. HOPPR is pioneering foundation models for medical imaging that have the potential to transform healthcare. With access to over 15 million diverse imaging studies, HOPPR is developing multimodal AI models that tackle radiology’s most significant challenges: high imaging volumes, limited specialist availability, and the growing demand for rapid, accurate diagnostics.

In this episode, Robert offers insight into the rigorous process of training these models on complex data while ensuring they integrate seamlessly into medical workflows. From data partnerships to specialized clinical collaboration, HOPPR’s approach sets new standards in healthcare AI. To discover how foundation models like these are revolutionizing radiology and making healthcare more efficient, accessible, and equitable, be sure to tune in today!


Key Points:

  • Robert’s background in medical imaging and tech and how it led him to create HOPPR.
  • Ways that HOPPR’s AI models improve diagnostic speed and accuracy.
  • The significant data and compute resources required to build a foundation model like this.
  • Partnering with imaging organizations to collect diverse data across multiple modalities.
  • How HOPPR differentiates itself with ISO-compliant development and multimodal training.
  • The quantitative metrics and clinical review involved in validating its foundation model.
  • Key challenges in building this model include data access, diversity, and secure handling.
  • Reasons that proper data diversity and balance are essential to reduce model bias.
  • How API integration makes HOPPR’s models easy to adopt into existing workflows.
  • The real-world clinical needs and input that go into building an AI product roadmap.
  • Robert’s take on what the future of foundation models for medical imaging looks like.
  • Valuable lessons on the importance of strong labeling, compute scalability, and more.
  • Practical, real-world advice for other leaders of AI-powered startups.
  • The broader impact in healthcare that HOPPR aims to make.


Quotes:

“Having clinical collaboration is super important. At HOPPR, our clinicians are an important part of our product development team – They're absolutely vital for helping us evaluate the performance of the model.” — Robert Bakos


“Because we are training across all these different modalities, getting access to this data can be challenging. Having great partnerships is critical for finding success in this space.” — Robert Bakos 


“Make sure that you're addressing real problems. There are a lot of great ideas and cool things you can implement with AI, but at the end of the day, you want to make sure you can deliver value to your customers.” — Robert Bakos


“Foundation models – trained on a breadth of data – can make a positive impact on underserved areas around the world. With the volume of images growing so rapidly, constraints on radiologists, and burnout, it's important to leverage these models to make a big impact.” — Robert Bakos


Links:

Robert Bakos

HOPPR

Robert Bakos on LinkedIn


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Smarter Farming with Eric Adamson from Tortuga AgTech07 Nov 202200:27:11

In this episode, I talk with Eric Adamson, CEO of Tortuga AgTech, about smarter farming. Tortuga AgTech builds robots for harvesting fruit and vegetables to help farms be more resilient, sustainable, and successful.


Quotes:


“Figuring out that pipeline from someone else's knowledge to the robot knows it is really critical.”


“If you build technology because the technology is cool or because you can, you are much more likely to fail than if you start with the customer problem and then figure out what kind of technology might help to solve that problem.”


“That learning happens with our machine learning engineers being in the field, being the ones who are actually taking data with handheld rigs.”


“Many of our team members’ first two weeks have been immediately flying to a farm and spending time on the farm with the robots, learning a problem in very, very deep detail. And I would encourage anybody building a technology based on machine learning or certainly robots to do the same.”


“We have a very efficient and effective pipeline that took us years to build. But it's exceptionally powerful for us to be able to, for example, go to a new site, run a couple robots or a small fleet of robots for a day, and then within a week have a brand new model that's been completely retrained on freshly labeled data from this new place.”


“That’s very critical for us because farm environments are changing so often. You really need to be able to be reactive and continue to improve your models as you develop.”


“We measure our scores based on golden data sets that we've sort of hand labeled ourselves. But we also have to make some judgment calls about what we really want in our performance versus what the conditions are in the field and what we're seeing on the farm.”


“We try to convert whatever model results are spit out into language that the customer intuitively understands.”


“It's really important to start with the customer problem and to start with the customer problem as an economic proposition.”


“There are already very large discussions happening in the farming community around what type of farming should be used in order to, for example, deal with climate change, to deal with drought, to deal with chemical regulations, to deal with a lowering of fruit quality and an increasing of fruit waste, the challenging labor environments.”


Links:

Tortuga AgTech

Twitter

YouTube

Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Computer Vision Advisory Services – Monthly advisory services to help you strategically plan your CV/ML capabilities, reduce the trial-and-error of model development, and get to market faster.

Autonomous Diagnostics with John Bertrand from Digital Diagnostics31 Oct 202200:21:33

In this episode, I talk with John Bertrand, CEO of Digital Diagnostics, about autonomous diagnostics. Digital Diagnostics transforms the quality, accessibility, equity, and affordability of healthcare with AI-powered diagnostics. They developed the first FDA-cleared autonomous AI system.


Quotes:

“So we look for diagnostics where there's an established understanding of what the disease is and there's a gold standard as to how to measure that.”


“We'll naturally start with an area where positive and negative is a very binary decision that is almost mathematically derived.”


“It goes back to picking the right types of disease states to make sure that the gold standard already exists.”


“How do you take images that have different coverage of the retina but make sure that you piece them together in a way that the processing part of the system is getting a consistent image that they're looking at every single time so that the algorithm remains consistent and we don't have to have different algorithms per vendor that we're interacting with.”


“We’re pretty proud of the fact we’ve been able to do that first kind of assistive feedback for the provider.”


“We want every single patient, regardless of their background, to receive consistent quality of diagnostic output. What that means is that we actually have to build our training data sets as well as our clinical validation studies and trials to take into account a diverse population set.”


“Continuous learning versus locked algorithms is another key factor. . . Would you really want that algorithm to adjust to the most recent data it's seeing, thinking it's attempting to become more accurate, when in fact it's really more optimizing for the ethnicity of the folks in that particular region, the sun rises on the east coast to the United States, everybody further east goes to bed. Now the algorithm’s been indexed towards another group from a ethnicity perspective, that’s no longer representative of where the testing’s being done as the sun rises in New York.”


“How do we ensure that we create confidence with regulators, with providers, and with patients that we've actually thought through this?”


“We can literally break down for you what the computer saw, why graded it out what it did, and why it gave you the results it did.”


“Your algorithm should be explainable so people trust the technology, understand how it works.”


“Also explainability helps you drive better accuracy and that you understand why you're getting the result that you're getting with the black box approach.”


“You really want to work within the healthcare system when you’re building these types of businesses.”


“If you're going to chart that course and really carry through to fruition, your vision of building an algorithm that impacts patient lives, I think you really need to center the culture of the business around a commonly shared vision for the mission of what you're trying to do.”


Links:

Digital Diagnostics

Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Computer Vision Advisory Services – Monthly advisory services to help you strategically plan your CV/ML capabilities, reduce the trial-and-error of model development, and get to market faster.

Capturing the Carbon Fingerprint of Soil with David Schurman from Perennial24 Oct 202200:23:18

In this episode, I talk with David Schurman, co-founder and CTO of Perennial, about their verification platform for climate-smart agriculture. Perennial uses geospatial data and machine learning to unlock agricultural soils as the world’s largest carbon sink.


Highlights:

  • How Perennial gathers and annotates training data from satellites and ground-based observations.
  • Handling variations across satellites and geographic locations.
  • Stratifying training data across the kinds of variables that matter.
  • Collaboration between machine learning engineers, remote sensing scientists, and crop scientists.
  • The importance of gathering more training data than you think you’ll need.
  • Respecting the data.
  • The nuance of communicating performance metrics.


Links:

Perennial’s website

Perennial on LinkedIn

David Schurman on LinkedIn

Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Computer Vision Advisory Services – Monthly advisory services to help you strategically plan your CV/ML capabilities, reduce the trial-and-error of model development, and get to market faster.

Biomarker Discovery from Pathology Images with Matt Alderdice from Sonrai Analytics17 Oct 202200:25:34

In this episode, I talk with Matt Alderdice, Head of Data Science at Sonrai Analytics, about precision medicine. Sonrai Analytics automates laborious data processes and speeds up new drug and healthcare developments.


Highlights:

  • Machine learning for automating time-consuming and tedious analysis of microscopy images.
  • Training for machine learning practitioners new to pathology by integrating domain experts with your team.
  • Involving stakeholders throughout a project.
  • Literature reviews to search for associated publications and potential solutions to avoid overly complicated solutions.
  • Validating models with ethnically diverse datasets.
  • Analytical validation for differing stains, scanners, and operators.
  • Clinical validation on a held out dataset in the same environment as would be in the clinic.
  • Identifying relevant metrics from conversations with pathologists, oncologists, nurses, and patients.
  • Focus on the problem you’re trying to solve – AI is just a tool.


Links:

Sonrai Analytics’ website

Matt Alderdice on LinkedIn

Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Computer Vision Advisory Services – Monthly advisory services to help you strategically plan your CV/ML capabilities, reduce the trial-and-error of model development, and get to market faster.

Diagnosing Emergent Diseases with David Golan from Viz.ai10 Oct 202200:32:03

In this episode, I talk with David Golan, co-founder and CTO of Viz.ai, about diagnosis of acute and emergent diseases. Viz.ai increases the speed of diagnosis and care for a variety of conditions to improve the lives of patients.


Highlights:

  • Increasing access to lifesaving treatments.
  • The importance of the full system, not just the machine learning component, in accelerating workflows.
  • Their clinical AI team includes med students, MDs, biomedical engineers, and neuropsychologists.
  • Bias can be created by a lower performance on a subset of the population in a way that is unknown to developers, users, and clinicians.
  • Careful monitoring of algorithms to identify subsets of data with poor performance.
  • Unbiased collection and stratification of data for FDA submission.
  • The importance of good annotation and monitoring infrastructure.
  • Relatively simple model architectures can take you a long way.


Links:

Viz.ai’s website

David Golan on LinkedIn

Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Computer Vision Advisory Services – Monthly advisory services to help you strategically plan your CV/ML capabilities, reduce the trial-and-error of model development, and get to market faster.

Environmental Risk Analysis with Steve Brumby from Impact Observatory10 Oct 202200:36:51

In this episode, I talk with Steve Brumby, co-founder, CEO and CTO of Impact Observatory, about sustainability and environmental risk analysis. Impact Observatory uses satellite imagery and machine learning to empower decision-makers with planetary insights.


Highlights:

  • Using machine learning to generate thematic maps to represent land cover and land use.
  • Geospatial data from the European Space Agency’s Copernicus program that is available on a variety of platforms.
  • The importance of identifying the relevant output for end users and others in the value chain.
  • How machine learning engineers sometimes discover things used by remote sensing scientists that are no longer necessary.
  • Keeping models simple.
  • Mitigating bias in models by using large and globally diverse datasets.
  • Get to know your customer and their pain points, then craft a machine learning solution that works for them – if you’re lucky, it’ll also work for others.
  • Finding the things you’re passionate about – both the technology and helping the customers in that space.


Links:

Resources for Computer Vision Teams:
LinkedIn – Connect with Heather.
Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.
Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.
Computer Vision Advisory Services – Monthly advisory services to help you strategically plan your CV/ML capabilities, reduce the trial-and-error of model development, and get to market faster.

Diagnosis and Management of Epilepsy with Dean Freestone from Seer10 Oct 202200:26:08

In this episode, I talk with Dean Freestone, co-founder and CEO of Seer, about epilepsy. Seer uses home monitoring to diagnose and manage neurological conditions, relieving bottlenecks in the healthcare system.


Highlights:

  • Using machine learning to summarize data to reduce the labor intensive search for episodic events.
  • Handling imbalanced datasets. 
  • Controlling the workflows to enable gathering and annotating huge datasets.
  • Working with technicians to speed up review of EEG data.
  • Using machine learning to capture features that doctors can’t describe.
  • Dealing with low inter-reviewer agreement from clinicians.
  • How bias can manifest is neurological data.
  • Do not underestimate the cost and amount of work to build a healthcare AI startup.


Links:

Seer’s website

Seer on LinkedIn

Dean Freestone on LinkedIn

Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Computer Vision Advisory Services – Monthly advisory services to help you strategically plan your CV/ML capabilities, reduce the trial-and-error of model development, and get to market faster.

Welcome to Impact AI05 Oct 202200:02:03

Welcome to Impact AI, the podcast for startups who want to create a better future through the use of machine learning.


I'm your host, Heather Couture.


In this podcast, you’ll learn how to build a mission-driven machine learning company. 


I’ll be interviewing innovators and entrepreneurs from a variety of industries: healthcare, drug development, environmental, agriculture, and many more.


Each is striving to solve a problem that they are passionate about. They will talk about the role machine learning plays in their technology and the impact of their product.


They will also help me uncover machine learning challenges like data annotation, generalizability, explainability, bias, and collaboration across disciplines – and best practices for tackling them in a startup environment.


Now, who am I?


I’m a consultant with almost 2 decades of experience in computer vision and machine learning for a variety of applications. From manufacturing to planetary science to commercial media to cancer research.


I completed a Masters at Carnegie Mellon University and a PhD in Computer Science at the University of North Carolina. As a researcher, I published in top-tier computer vision and medical imaging venues. Now I write regularly on LinkedIn, for my newsletter Pathology ML Insights, and for a variety of trade publications.


I offer consulting services through my company Pixel Scientia Labs to help startups get to market faster by building more generalizable computer vision models. I make use of the latest machine learning research to amplify their results and support their in-house team for the long term. My mission is to fight cancer and climate change with AI – and I do that by strengthening the machine learning component of my clients’ most impactful projects.


My hope for this podcast is to share machine learning best practices more widely so that many others can benefit as they work towards solving important problems.


Thanks for listening.


Please hit subscribe to be notified about new episodes.

Optimizing Data Center Operations with Vedavyas Panneershelvam from Phaidra21 Oct 202400:22:14

What are the unique challenges of operating mission-critical facilities, and how can reinforcement learning be applied to optimize data center operations? In this episode, I sit down with Vedavyas Panneershelvam, CTO and co-founder of Phaidra, to discuss how their cutting-edge AI technology is transforming the efficiency and reliability of data centers. Phaidra is an AI company that specializes in providing intelligent control systems for industrial facilities to optimize performance and efficiency. Vedavyas is a technology entrepreneur with a strong background in artificial intelligence and its applications in industrial and operational settings. In our conversation, we discuss how Phaidra’s closed-loop, self-learning autonomous control system optimizes cooling for data centers and why reinforcement learning is the key to creating intelligent systems that learn and adapt over time. Vedavyas also explains the intricacies of working with operational data, the importance of understanding the physics behind machine learning models, and the long-term impact of Phaidra’s technology on energy efficiency and sustainability. Join us as we explore how AI can solve complex problems in industry and learn how Phaidra is paving the way for the future of autonomous control with Vedavyas Panneershelvam.


Key Points:

  • Hear how collaborating on data center optimization at Google led to the founding of Phaidra.
  • How Phaidra’s AI-based autonomous control system optimizes data centers in real-time.
  • Discover how reinforcement learning is leveraged to improve data center operations.
  • Explore the range of data needed to continuously optimize the performance of data centers.
  • The challenges of using real-world data and the advantages of redundant data sources. 
  • He explains how Phaidra ensures its models remain accurate even as conditions change.
  • Uncover Phaidra’s approach to validation and incorporating scalability across facilities. 
  • Vedavyas shares why he thinks this type of technology is valuable and needed.
  • Recommendations for leaders of AI-powered startups and the future impact of Phaidra.


Quotes:

“Phaidra is like a closed-loop self-learning autonomous control system that learns from its own experience.” — Vedavyas Panneershelvam


“Data centers basically generate so much heat, and they need to be cooled, and that takes a lot of energy, and also, the constraints in that use case are very, very narrow and tight.” — Vedavyas Panneershelvam


“The trick [to validation] is finding the right balance between relying on the physics and then how much do you trust the data.” — Vedavyas Panneershelvam


“[Large Language Models] have done a favor for us in helping the common public understand the potential of these, of machine learning in general.” — Vedavyas Panneershelvam


Links:

Vedavyas Panneershelvam on LinkedIn

Phaidra


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Structuring Medical Text with Tim O'Connell from Emtelligent14 Oct 202400:18:02

What if AI could unlock the potential of healthcare’s vast, unstructured data? In this episode, Tim O'Connell, Co-Founder and CEO of Emtelligent, explains how his company is bridging the gap between messy medical data and usable insights with AI-powered solutions. Drawing from his background in both engineering and radiology, Tim discusses how he saw firsthand the inefficiencies caused by disorganized medical notes and reports, which led to the creation of Emtelligent. He breaks down how their AI models work to process and structure this data, making it usable for healthcare professionals, researchers, and beyond. Tim also dives into the technical challenges, from handling faxed medical records to ensuring high levels of precision and recall in model training. Beyond the technology, he emphasizes the importance of safety, ethical use, and how Emtelligent continues to adapt its AI to meet the evolving needs of the healthcare industry, helping to make patient care more efficient and accurate. Don’t miss out on this important conversation with Tim O’Connell from Emtelligent!


Key Points:

  • An overview of Tim’s background in engineering and radiology.
  • How Tim co-founded Emtelligent to solve pressing data issues in healthcare.
  • The importance of turning unstructured medical text into searchable, structured data.
  • How Emtelligent’s models extract metadata and structure from faxed patient records.
  • Why healthcare data is so challenging to work with, from shorthand to messy notes.
  • The role of precision and recall in assessing and improving model performance in healthcare.
  • Ensuring AI models continue to perform well after deployment with ongoing updates.
  • How Tim’s team maintains safety and ethical standards in AI healthcare solutions.
  • Creating technology that serves the end user; how it is informed by firsthand experience.
  • The importance of clinical input to develop relevant and practical AI healthcare tools.
  • Where Tim sees AI's impact in healthcare evolving over the next three to five years.


Quotes:

“During that year [that I was] working in the hospital, – I saw so many problems that we have in the healthcare environment and realized that quite a few of them had to do with the fact [that] we deal with so much unstructured data.” — Tim O’Connell


“Every time a human goes to see a caregiver, some kind of an unstructured text note is generated – We really can't use a lot of that data, unless it's another human who's reading that data.” — Tim O’Connell


“I’m still a practicing radiologist. – It’s not just a matter of intelligent people coming up with good ideas and going, ‘Oh, well. [Let’s throw this] against the wall and see what sticks’. We're developing solutions that are applicable in today's healthcare environment.” — Tim O’Connell


Links:

Tim O’Connell on LinkedIn

Emtelligent


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Foundation Model Series: Enabling Digital Pathology Workflows with Dmitry Nechaev from HistAI07 Oct 202400:29:32

What happens when you combine AI with digital pathology? In this episode, Dmitry Nechaev, Chief AI Scientist and co-founder of HistAI, joins me to discuss the complexity of building foundation models specifically for digital pathology. Dmitry has a strong background in machine learning and experience in high-resolution image analysis. At HistAI, he leads the development of cutting-edge AI models tailored for pathology.

HistAI, a digital pathology company, focuses on developing AI-driven solutions that assist pathologists in analyzing complex tissue samples faster and more accurately. In our conversation, we unpack the development and application of foundation models for digital pathology. Dmitry explains why conventional models trained on natural images often struggle with pathology data and how HistAI’s models address this gap. Learn about the technical challenges of training these models and the steps for managing massive datasets, selecting the correct training methods, and optimizing for high-speed performance. Join me and explore how AI is transforming digital pathology workflows with Dmitry Nechaev!


Key Points:

  • Background about Dmitry, his path to HistAI, and his role at the company.
  • What whole slide images are and the challenges of working with them.
  • How AI can streamline diagnostics and reduce the workload for pathologists.
  • Why foundation models are a core component of HistAI’s technology. 
  • The scale of data and compute power required to build foundation models.
  • Outline of the different approaches to building a foundation model.
  • Privacy aspects of building models based on medical data.
  • Challenges Dmitry has faced developing HistAI’s foundation model. 
  • Hear what makes HistAI’s foundation model different from other models.
  • Learn about his approach to benchmarking and improving a model. 
  • Explore how foundation models are leveraged in HistAI’s technology. 
  • The future of foundation models and his lessons from developing them.
  • Final takeaways and how to access HistAI’s open-source models.


Quotes:

“Regular foundation models are trained on natural images and I'd say they are not good at generalizing to pathological data.” — Dmitry Nechaev


“In short, [a foundational model] requires a lot of data and a lot of [compute power].” — Dmitry Nechaev

“Public benchmarks [are] a really good thing.” — Dmitry Nechaev


“Our foundation models are fully open-source. We don't really try to sell them. In a sense, they are kind of useless by themselves, since you need to train something on top of them, so we don't try to profit from these models.” — Dmitry Nechaev


“The best lesson is that you need quality data to get a quality model.” — Dmitry Nechaev


“[HistAI] don't want AI technologies to be a privilege of the richest countries. We want that to be available around the world.” — Dmitry Nechaev


Links:

Dmitry Nechaev on LinkedIn

Dmitry Nechaev on GitHub

HistAI

CELLDX

Hibou on Hugging Face


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Foundation Model Series: Creating Small Molecules for Drug Discovery with Jason Rolfe from Variational AI30 Sep 202400:29:29

Building on the trends in language processing, domain-specific foundation models are unlocking new possibilities. In the realm of drug discovery, Jason Rolfe is spearheading innovation at the intersection of AI and pharmaceuticals. As the Co-Founder and CTO of Variational AI, Jason leads a platform designed to generate novel small molecule structures that accelerate drug development. In this episode, he delves into how Variational AI uses foundation models to predict and optimize small molecules, overcoming the immense complexity of drug discovery by leveraging vast datasets and sophisticated computational techniques. He also addresses the key challenges of modeling molecular potency and why traditional machine-learning approaches often fall short. For anyone curious about AI's impact on healthcare, this conversation offers a fascinating look into cutting-edge innovations set to reshape the pharmaceutical industry. Tune in to find out how the types of breakthroughs we discuss in this episode could revolutionize drug development, bring new therapeutics to market across disease areas, and positively impact lives!


Key Points:

  • An overview of Jason’s background and how it led him to create Variational AI.
  • What Variational AI does for the small molecule domain for drug discovery.
  • How they use foundation models to predict and enhance the design of small molecules.
  • Defining small molecules, their appeal, and an overview of Variational AI's data sets.
  • What goes into training Variational AI's foundation model.
  • The computational infrastructure and algorithms necessary to process this data.
  • Challenges of predicting molecular potency against disease-related protein targets.
  • Various ways that Variational AI’s foundation model underpins everything they do.
  • Evaluating progress: balancing predictive success with experimental validation.
  • Lessons from developing foundation models that could apply to other data types.
  • Jason’s funding and research-focused advice for leaders of AI-powered startups.
  • The transformative impact of Variational AI’s technology on drug development.


Quotes:

“Rather than forming individual models for specific drug targets, we're creating a joint model over hundreds, eventually thousands of drug targets.” — Jason Rolfe


“Data quality is essential. In particular, if you're drawing from multiple different data sources, frequently, those sources aren't commensurable.” — Jason Rolfe


“If you don't have a proven track record where people are already throwing money at you, it is very challenging to try to bring a new technology from the drawing board into commercial application using venture funding.” — Jason Rolfe


“Whenever you're developing a new technology or product, you need to test early and often. Some of your intuitions will be good. Most of your intuitions will be a waste of time – The more quickly you can distinguish between those two classes, the more efficiently you can move toward success.” — Jason Rolfe


Links:

Variational AI

Variational AI Blog

Jason Rolfe on LinkedIn


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Foundation Model Series: Building New Materials for Climate with Jonathan Godwin from Orbital Materials23 Sep 202400:25:03

AI is unlocking the future of materials science and today’s guest Jonathan Godwin, co-founder and CEO of Orbital Materials, is at the forefront of this transformation. With a background in AI research and experience leading groundbreaking projects at Google-owned DeepMind, Jonathan is now applying machine learning to develop advanced materials that can drive decarbonization.

In this episode, he explains how Orbital Materials is using foundation models (like ChatGPT for language or MidJourney for images) to design new materials that capture carbon, store energy, and improve industrial efficiency. He also shares insights into the company’s mission, the challenges of simulating atomic-level interactions, and why open-sourcing their model, Orb, is crucial for innovation.

To discover how AI is revolutionizing the fight against climate change and learn how these cutting-edge materials could shape a more sustainable future, don’t miss this inspiring conversation with Jonathan Godwin!


Key Points:

  • Insight into Jonathan’s diverse career path and how it led him to Orbital Materials.
  • What types of advanced materials Orbital develops and their potential impact.
  • The critical role AI plays in developing materials for decarbonization purposes.
  • Defining foundation models and why they’re an essential part of leveraging AI.
  • 3D atomic simulations and other types of data that go into Orbital’s foundation model.
  • The computing infrastructure required to build a foundation model for materials.
  • Engineering and other challenges encountered while building models at this scale. 
  • How AI enhances scientific discovery without replacing human expertise.
  • Why open-sourcing Orbital’s foundation model, Orb, is key for innovation.
  • Lessons from developing this model that could be applied to other data types.
  • Jonathan’s detail-oriented advice for leaders of AI-powered startups.
  • Orbital’s exciting mission to accelerate new materials development.


Quotes:

“We develop materials that can capture CO2 from specific gas streams – coming out of an industrial facility, new energy storage technologies that allow – [data centers] to operate behind the meter, or ways to improve the water efficiency of a data center or industrial facility.” — Jonathan Godwin


“Foundation models are the crux of how we're able to leverage AI in this day and age. If you want to [say], 'We're pushing the limits of what AI is able to do. We're leveraging the most recent breakthroughs,' – you've got to be building foundation models or using foundation models.” — Jonathan Godwin


“AI is a massively powerful creativity aid and accelerant. We’ve seen that in other areas of AI and we're bringing that to advanced materials.” — Jonathan Godwin


Links:

Orbital Materials

Orbital Materials on LinkedIn

Orbital Materials on X

Orbital Materials on GitHub

Jonathan Godwin on LinkedIn

Jonathan Godwin on X

Jonathan Godwin Substack


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Foundation Model Series: Understanding Brain Activity with Dimitris Sakellariou from Piramidal16 Sep 202400:23:41

What if we could understand brain activity in real-time to better diagnose neurological conditions? In this episode, part of a special mini-series on domain-specific foundation models, I sit down with Dimitris Sakellariou, the founder and CEO of Piramidal, to talk about their groundbreaking work in automating EEG interpretation. Piramidal is focused on democratizing brain health insights, making interpreting brainwave data more accessible and accurate. With a strong foundation in neuroscience and AI, Dimitris and his team are developing models that could revolutionize how we understand brain activity and diagnose neurological conditions.

In our conversation, Dimitris explains the challenges of building a foundation model for brain activity, the role of data diversity, and the future potential for personalized brain health monitoring. Discover the implications of Piramidal’s technology beyond healthcare and its application in cognitive enhancement and stress management. Tune in as we explore how Piramidal is paving the way for personalized brain health monitoring and why this could be a game-changer for the future of medicine!


Key Points:

  • Dimitris discusses his journey from physics to a career in neuroscience.
  • Explore Piramidal's mission to automate EEG interpretation.
  • Learn about the complexity and variability of brainwave patterns
  • Hear how machine learning can better analyze brain activity.
  • Uncover the challenges of building a foundation model for EEG data.
  • Why diverse data sets are vital for training the foundational model.
  • Piramidal's plans for making EEG analysis more accessible.
  • Future use cases for Piramidal’s model in healthcare and beyond.
  • Discover why domain knowledge for model building is essential.
  • He shares advice for AI startup founders.


Quotes:

“Piramidal is primarily focused at the moment in automating, or otherwise democratizing the interpretation of these tests, these brainwave recordings so that patients and people that have issues with their brain can get access to the diagnosis much, much, much faster.” — Dimitris Sakellariou

“It's very important to have discussions with neuroscientists and clinical experts in order to understand what is the end-to-end pipeline from receiving data all the way to inference.” — Dimitris Sakellariou


“Finding the right person. Someone that is very keen to build together with you and make important and difficult decisions can change massively a trajectory of your company.” — Dimitris Sakellariou


Links:

Dimitris Sakellariou on LinkedIn

Dimitris Sakellariou on X

Piramidal

Piramidal on LinkedIn


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Foundation Model Series: Better, Faster, Cheaper Earth Observation with Bruno Sánchez-Andrade Nuño from Clay09 Sep 202400:35:35

Can AI be applied to enhance geospatial data for climate, nature and people? This episode kicks off a miniseries about domain-specific foundation models. Following the trends in language processing, domain-specific foundation models are enabling new possibilities for a variety of applications, including Earth observation. During this conversation, I am joined by Bruno Sánchez-Andrade Nuño, Executive Director of Clay, a nonprofit organization harnessing the power of AI for satellite images, spatial data, and more. Bruno shares the functionality and concept behind Clay, and his journey to building it. He goes on to unpack the tool’s foundation model in broad strokes, before explaining why it's important, and sharing the challenges he has faced along the way. We discuss the legal aspects of building Clay, and it’s primary goal to make it as easy as possible for any user to achieve their goals. We also touch on what the future might hold for Clay and the future of Earth observation. Thanks for listening!


Key Points:

  • Introducing guest, Bruno Sánchez-Andrade Nuño, Executive Director at Clay.
  • His journey from NASA astrophysicist to climate change, social development, and AI researcher.
  • What Clay focuses on: using remote sensing maps to interpret the Earth’s data.
  • The mechanics of how Clay is used and how different feature sets compare to one another.
  • A broad explanation of the tool’s foundation model and why it is quicker, cheaper, and more environmentally friendly.
  • Two main benefits of the tool that Bruno finds most exciting. 
  • Data and infrastructure required to build Clay including 70 million satellite and aerial images.
  • Measuring what the model understands and the process of compressing an image into 700 numbers.
  • Privacy and intellectual property in the realm of satellite imaging and mapping. 
  • What commercial imagery could add to the model and how it might be integrated in the future. 
  • Clay’s partnerships with university and company groups
  • Why the focus of Clay is to make it as easy as possible for anyone to use the tool for anything they want to do. 
  • Challenges encountered on the road to building Clay: explaining what it is.
  • The complexity of benchmarking foundation models and how this relates to Clay. 
  • Working with partners to build Clay and the rest of the ecosystem. 
  • Lessons from building Clay that may apply to other foundation models.
  • Bruno’s predictions for the future of foundation models and Clay. 
  • What is certain about the future of Clay and our understanding of Earth. 


Quotes:

“Clay is trying to figure out how to finally increase the adoption of remote sensing by leveraging a tool that itself is very complex, but the result of that tool is very easy to use.” — Bruno Sánchez-Andrade Nuño


“If you start with a foundational model that gets you most of the way there, [then] you can create those trials much quicker, much cheaper, and much more environmentally friendly.” — Bruno Sánchez-Andrade Nuño


“This is so new, we get the chance, those of us working on it, that we can save the whole industry, if you will, the whole space of AI for it.” — Bruno Sánchez-Andrade Nuño


“Clay, I believe, is not only the largest and most efficient model AI for Earth, for any kind of like foundational model. It is also completely open source.” — Bruno Sánchez-Andrade Nuño


“What we try to focus on is how can we make it as simple as possible for anyone anywhere to use this model for anything they want to do.” — Bruno Sánchez-Andrade Nuño


Links:

Bruno Sánchez-Andrade

Bruno Sánchez-Andrade Nuño on X

Bruno Sánchez-Andrade Nuño on LinkedIn

Clay

Clay on LinkedIn


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Evolutionary Insights for Drug Discovery with Ashley Zehnder from Fauna Bio02 Sep 202400:27:17

In a world where conventional drug discovery methods frequently fall short, today's guest addresses the critical challenge of fighting human diseases by drawing inspiration from nature’s most resilient creatures. Could the secret to overcoming our most stubborn illnesses lie in the extraordinary adaptability of extreme mammals? Veterinarian-scientist Ashley Zehnder, the Co-founder and CEO of AI-driven drug discovery company Fauna Bio, believes so.

By leveraging data from 100 million years of evolved disease resistance in mammals, Ashley sees a unique opportunity at the crossroads of genomics and emerging model species to improve health for all species, including humans. In this episode, she explores how harnessing the biological secrets of these animals using AI and machine learning could revolutionize medicine, leading to breakthroughs that benefit us all. Tune in to discover how Fauna Bio is pioneering a new frontier in drug discovery and how understanding the resilience of these creatures could reshape the future of healthcare!


Key Points:

  • Insight into the diverse backgrounds of Fauna Bio’s founding members.
  • Ways that Fauna Bio uses AI and genomics to identify key targets for new therapeutics.
  • The role machine learning plays in analyzing and annotating large volumes of data.
  • Gene expression and other data inputs that drive Fauna Bio’s discoveries.
  • The collaborative effort required to collate datasets from 400+ mammals.
  • Challenges of working with genomic data and training ML models on it.
  • How Fauna Bio rigorously validates their AI-driven discoveries.
  • Cooperation between ML developers and domain experts to advance this technology.
  • Technological advancements that enable Fauna Bio’s innovations.
  • Ashely’s advice on differentiation for leaders of AI-powered startups.
  • Where she sees Fauna Bio making the biggest impact in the future.


Quotes:

“[Fauna Bio uses] AI and genomics as a way to identify the most impactful targets for new therapeutic programs across a broad number of diseases.” — Ashley Zehnder


“It’s certainly easier than it has been in the past to generate very high-quality single-cell RNA sequencing. We’re doing a lot of that. The challenges on the technical side are getting much easier. The challenges on the interpretation side are still there.” — Ashley Zehnder


“There are many points along the drug discovery path where AI companies can differentiate. But that story has to be clear because, otherwise, it's very hard to get out of the signal-to-noise that is the AI discovery landscape in biopharma” — Ashley Zehnder


Links:

Fauna Bio

Ashley Zehnder on LinkedIn

Ashley Zehnder on X

Ashley Zehnder Email

Zoonomia Project

Science Issue dedicated to the Zoonomia Project


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Foundation Model Series: Harnessing Multimodal Data to Advance Immunotherapies with Ron Alfa from Noetik03 Feb 202500:33:53

In this episode, I'm joined by Ron Alfa, Co-Founder and CEO of Noetik, to discuss the groundbreaking role of foundation models in advancing cancer immunotherapy. Together, we explore why these models are essential to his work, what it takes to build a model that understands biology, and how Noetik is creating and sourcing their datasets. Ron also shares insights on scaling and training these models, the challenges his team has faced, and how effective analysis helps determine a model’s quality. To learn more about Noetik’s innovative achievements, Ron’s advice for leaders in AI-powered startups, and much more, be sure to tune in!

Key Points:

  • Ron shares his background and how his journey led to Noetik.
  • Why a foundation model is important in their work.
  • What goes into building a foundation model that understands biology.
  • Building the dataset: where does the data come from?
  • The types of data they generate from the samples they use in their models.
  • He further explains the components necessary to build a foundation model.
  • The scale and what it takes to train these models. 
  • Ron sheds light on the challenges they’ve encountered in building their foundation model.
  • How to determine if your foundation model is good. 
  • Utilizing analysis to help identify ways to improve your model. 
  • The current purpose for their foundation model and how they plan to use it in the future.
  • Key insights gained from developing foundation models and how these can be adapted to other types of data.
  • His advice to other leaders of AI-powered startups.
  • Ron digs deeper into their goal to impact patient care by developing new therapeutics.


Quotes:

“Our thesis for Noetik is that one of the biggest problems we can impact if we want to make and bring new drugs to patients is predicting clinical success; so-called translation — that's where we focus Noetik, how can we train foundation models of biology so that we can better translate therapeutics from early discovery and preclinical models to patients.” — Ron Alfa


“We think the most important thing for any application of machine learning is the data.” — Ron Alfa


“The goal here is to train models that can do what humans cannot do, that can understand biology that we haven't discovered yet.” — Ron Alfa


“The big aim of Noetik is to develop these [foundational] models for therapeutics discovery.” — Ron Alfa


Links:

Ron Alfa on LinkedIn

Ron Alfa on X

Noetik

Noetik Octo Virtual Cell (OTCO)


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Better Therapeutics Using Lab-Grown Tissue with Andrei Georgescu from Vivodyne26 Aug 202400:33:57

One of the biggest hurdles in medical research is the gap between animal studies and human trials, a disconnect that often leads to failed drug tests and wasted resources. But what if there was a way to bridge that gap and create treatments that are more effective for humans from the start?

Today, I am joined by Dr. Andrei Georgescu, Founder and CEO of Vivodyne, a groundbreaking biotechnology company that is transforming how scientists study human biology and develop new therapeutics. In this episode, he reveals how Vivodyne harnesses lab-grown tissue and advanced multimodal AI to create more effective therapeutics. We explore the challenges of gathering human tissue data, the collaboration between biologists, robotics engineers, and machine learning developers to build powerful machine learning models, and the profound impact that Vivodyne is poised to make in the fight against diseases. To discover how Vivodyne’s innovations can lead to more successful treatments and faster drug development, tune in today!


Key Points:

  • Insight into Andrei’s background and how it led him to create Vivodyne.
  • What Vivodyne does and why it’s so important for drug discovery.
  • The role that AI and machine learning play in analyzing vast amounts of data.
  • Different data inputs and outputs for Vivodyne’s advanced multimodal AI.
  • The value of biased and unbiased AI outputs depending on the context.
  • Why interpretability and explainability are crucial in fields like biotechnology.
  • Challenges associated with collecting human tissue data to train Vivodyne’s models.
  • What goes into validating Vivodyne’s machine learning models.
  • Difficulties in integrating biology knowledge with robotics and machine learning.
  • Andrei’s business-focused advice for technical founders.
  • The profound impact that Vivodyne will have on drug discovery in the future.


Quotes:

“Vivodyne grows human tissues at a very large scale so that we can understand human physiology and we can test directly on it in order to discover and develop better drugs that are both safer and more efficacious.” — Andrei Georgescu


“We use machine learning and AI as a mechanism to understand the complexity of very deep data and to very efficiently apply that complexity and infer from what we've learned across the very large breadth of data that we collect.” — Andrei Georgescu


“To address [the problem of a] glaring lack of trainable data, we create that data by growing it at scale.” — Andrei Georgescu


“If you're a technical founder, do something that is incredibly hard because the ability for you to do that thing will grant you much more leverage than creating what is otherwise a much more simple and generic business.” — Andrei Georgescu


“[With Vivodyne], we will enter a world of plenty where the development of new drugs against diseases becomes a far more successful, reliable, and predictive process, and we're able to make much safer and much more effective drugs just by virtue of being able to optimize that therapeutic on human tissues before giving it to people for the first time in-clinic.” — Andrei Georgescu


Links:

Andrei Georgescu

Vivodyne

Andrei Georgescu on LinkedIn


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Accelerating Regenerative Agriculture with Marie Coffin from CIBO Technologies19 Aug 202400:16:02

Marie Coffin is the Vice President of Science and Modeling at CIBO Technologies, and she is with me today to discuss regenerative agriculture. Join us as we explore CIBO’s work to influence company carbon footprints across industries, and how machine learning supports this process through remote sensing. Delving deeper, Marie unpacks how satellite imagery integrates with their computer vision system for a more scalable solution. Next, we discuss obtaining and categorizing data in the US, exploring some of the obstacles that stem from privacy and data protection concerns. We touch on data quality and discuss the reason behind the geographical parameters they have applied to the work before Marie shares her approach to collaborating with external experts and agronomists. She offers her advice for startups in the tech space, emphasizing creating value for your clients over keeping up with trends, predicts the future endeavors that CIBO will focus on, and more. Thanks for listening! 


Key Points:

  • Introducing Marie Coffin and her background leading up to her role at CIBO Technologies.
  • CIBO’s work to influence company carbon footprints to improve agricultural sustainability.
  • The role of machine learning in this process: remote sensing.
  • What remote sensing is used for at CIBO.
  • How satellite imagery interacts with their computer vision system. 
  • Gathering, labeling, and annotating data with a focus on the boundary of the field. 
  • Obtaining this information through a farmer’s recording process. 
  • Why their work is largely limited to the US at the moment. 
  • Challenges related to privacy and data protection while working with training models.
  • Managing data quality issues.
  • Validating models within a geographical context. 
  • Collaborating with domain experts and external agronomists to understand and validate thier approaches.
  • How the seasonal nature of agriculture impacts the timing of reports and outputs. 
  • Advice for tech startups; addressing trends, who to hire, and more.
  • Qualities Marie seeks in new hires. 
  • Her prediction for CIBO’s growing impact in the next three to five years. 


Quotes:

“It’s pretty straightforward to estimate the carbon footprint of a single farmer’s field or even the carbon footprint of a whole farm, but, to make an impact, we need to be able to scale that across the landscape.” — Marie Coffin


“That is really the biggest challenge; it’s just getting enough data.” — Marie Coffin


“When you’re working in a really cutting-edge area, it’s tempting to sort of get caught up in the buzz of the new technology and lose sight of what the customer needs.” — Marie Coffin


“We need to not always be following the latest, greatest advance. We need to be going in a direction that’s going to really provide value.” — Marie Coffin


Links:

CIBO Technologies

Marie Coffin on LinkedIn


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Measuring Biodiversity Using Insects with Mads Fogtmann from Fauna Photonics12 Aug 202400:20:34

What if technology could be the key to averting a biodiversity crisis? Today, I explore this possibility with Mads Fogtmann, Chief Data Officer of FaunaPhotonics, as we discuss their groundbreaking approach to biodiversity monitoring. I talk with Mads about the looming biodiversity crisis, the innovative solutions his team is developing to address the urgent need for scalable biodiversity monitoring, and the central role that humans have to play in all this. Find out how the FaunaPhotonics platform is employing advanced sensing technology and machine learning to protect ecosystems, why insects are such useful proxies for monitoring ecosystem health, and their successful partnerships with other domain experts and researchers. Our conversation also covers the broader implications of biodiversity loss, the role of public awareness in conservation, and the future of biodiversity monitoring. Join us for a comprehensive and insightful discussion on how technology can help safeguard our planet's future and ensure the stability of natural and human systems alike!


Key Points:

  • Some background on Mads and his transition from academia to the private sector.
  • The FaunaPhotonics platform and how it monitors biodiversity.
  • An overview of the biodiversity crisis and the urgent need to address it.
  • Understanding our connection to, and dependence on, nature.
  • The risks that the biodiversity crisis poses for supply chains.
  • FaunaPhotonics’ role in measuring the biodiversity crisis: why this protects ecosystems.
  • Why insects are the best available proxy for measuring ecosystem health.
  • How sensing technology and machine learning are utilized by FaunaPhotonics.
  • Case studies showcasing the impact of FaunaPhotonics' technology.
  • Future directions and innovations in biodiversity monitoring.
  • Key challenges faced in developing and deploying biodiversity monitoring technology.
  • FaunaPhotonics’ collaboration with other domain experts and researchers in the field.
  • Why there is such an urgent need for scaleable biodiversity monitoring.
  • The importance of public awareness and education in addressing the biodiversity crisis.
  • Mads’ advice to leaders of other AI-powered startups and the future of FaunaPhotonics.


Quotes:


“The clothes we wear, the food we eat, the water we drink, the material we use to build houses: everything comes from nature. And right now, we are destroying that foundation rapidly.” — Mads Fogtmann


“I think it’s important that we become more aware that we are an integral part of nature.” — Mads Fogtmann


“If you can’t measure it, then how can you protect the rights? – [We come with the solution] that allows them to measure [the impact on biodiversity] so they can protect it. We do this by using insect sensing. The reason we do this is that insects are so fundamental to the ecosystem.” — Mads Fogtmann

“Insects are the best proxy that you can have for actually measuring the health of [an] ecosystem.” — Mads Fogtmann


“There’s a huge need and an interest in ‘how we can actually scale biodiversity monitoring to kind of help us understand what’s going on with nature at the moment.’” — Mads Fogtmann


Links:

Mads Fogtmann on LinkedIn
FaunaPhotonics

FaunaPhotonics on LinkedIn


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Optimizing Manufacturing with Berk Birand from Fero Labs05 Aug 202400:20:50

Manufacturing is a fundamental part of our economy. Unfortunately, a huge swath of the industry is still dependent on outdated methods, adversely impacting our environment. To address these challenges, one company is harnessing the power of AI to transform traditional manufacturing, driving unprecedented efficiency and sustainability in the industry. Joining me today is Berk Birand, co-founder and CEO of Fero Labs, to unpack how AI is optimizing the manufacturing sector.

Tuning in, you'll learn all about Fero Labs' innovative software and how it’s empowering engineers in industries like steel and chemicals to harness machine learning, drastically reducing waste and energy consumption. We discuss how their AI analyzes historical production data to ensure factories operate at peak performance and how this is boosting sustainability and profitability. Our conversation also unpacks the critical role of explainable AI in building trust within the industrial sector, where precision and reliability are essential. Tune in to discover how Fero Labs is paving the way for a greener industrial future!


Key Points:

  • Berk Birand’s education and career background.
  • How he co-founded Fero Labs with his business partner.
  • An overview of Fero Labs’ AI software.
  • Fero Labs’ role in reducing raw material waste in the steel industry.
  • How they have helped improve energy efficiency in chemical manufacturing.
  • Data analysis and how their software provides recommendations for efficient operations.
  • Understanding the high stakes involved in manufacturing processes.
  • Why AI explainability is crucial in the industrial sector.
  • How they are building explainable models that engineers can trust and understand.
  • Why now is the right time to build this technology.
  • His advice to AI-powered startups: seriously consider the cost of a bad prediction.
  • Fero Labs’ long-term vision to achieve a more circular and sustainable industrial sector.


Quotes:

"One of our largest customers was able to reduce the waste of raw materials, about a million pounds just throughout last year, by using our software AI system." — Berk Birand


"We think AI will play a key role in the transition to a green economy." — Berk Birand


"The best people to be solving these types of challenges, ultimately, are the engineers that work at the plants. The engineers that have the most domain expertise." — Berk Birand


"In an environment like this, an engineer in a factory would just not want to use a software that they don't trust, because ultimately, it's their job that's on the line." — Berk Birand


“With the new drive towards building an industrial sector that is more circular and more sustainable, there's incredible potential to optimize not just an individual factory, but beyond that, to optimize the entire supply chain by optimizing factories jointly.” — Berk Birand

Links:

Berk Birand on LinkedIn

Fero Labs


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

More Successful IVF with Daniella Gilboa from AIVF29 Jul 202400:27:38

In this episode of Impact AI, we delve into the transformative impact of AI on in-vitro fertilization (IVF) with Daniella Gilboa, co-founder and CEO of AIVF, a startup that develops AI-powered IVF solutions to help increase the certainty of a successful journey to parenthood. Join me as Daniella shares her mission to democratize fertility care and offers insight into AIVF’s proprietary technology that delivers reliable, objective, and data-driven IVF outcomes for clinicians, embryologists, and patients. We explore the role and challenges of machine learning at AIVF, strategies for validating AI models in clinical practice, and the current demand for AI-powered IVF solutions. We also discuss the metrics used to measure the impact of AIVF's technology, Daniella’s advice for other AI-powered startup leaders, and her vision for the future. Tune in to gain valuable insights into the future of fertility care and find out how AI is making IVF more effective and accessible!


Key Points:

  • How Daniella came to understand the epidemiology and data aspects of fertility.
  • What AIVF does and why it’s so important for both patients and clinicians.
  • The role of machine learning at AIVF and the challenges their models encounter. 
  • AIVF’s strategy for validating their models and translating KPIs into clinical settings.
  • The value of explainability to empower embryologists to use AI as a tool.
  • Daniella’s definition of computational embryology, assisted by machine learning.
  • Why now is the right time for AI-powered IVF solutions.
  • Metrics that AIVF uses to measure the impact of their technology.
  • Danielle’s advice for leaders of AI-powered startups and her vision for the future.


Quotes:

“We showed that if you use AI as a tool for the embryologist – [it] increased the success rates – The decision-making is faster, more accurate. You freeze less embryos because each embryo you freeze is accurate – It changes the way the lab works and it optimizes everything.” — Daniella Gilboa


“The way you interact with the patient and consult the journey ahead is changing. It’s more accurate. It allows you to make more informed decisions. This is the right way of doing medicine. It needs to be data-driven rather than subjective human analysis.” — Daniella Gilboa


“AIVF needs to become the standard of care.” — Daniella Gilboa


Links:

AIVF

Daniella Gilboa on LinkedIn

Daniella Gilboa on X


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Vision Intelligence Filters with Kit Merker from Plainsight Technologies22 Jul 202400:28:13

Image-based machine learning is fast becoming an AI staple, and with its new Vision Intelligence Filters, Plainsight Technologies is staking its claim as an industry pioneer. Today, I am joined by Plainsight CEO, Kit Merker, who is here to share all the details behind his company’s latest innovation. Kit begins by explaining what Plainsight does and why this work matters in the AI realm. Then, we learn about the mechanics behind Plainsight’s Vision Intelligence Filters, the company’s ML models and data protocols concerning existing customers, the ins and outs of bringing a product like the Vision Intelligence Filters to life, and how bias manifests in image-trained models. We also discuss the most game-changing applications that Kit has been involved in, and he shares some critical advice for young leaders of AI-powered startups, plus so much more!


Key Points:

  • Kit’s professional background and how he ended up at Plainsight.
  • What Plainsight does and why this work matters. 
  • The mechanics behind Plainsight's Vision Intelligence Filters.
  • How the company's ML models and data use relate to its customers 
  • Understanding when domain expertise comes into play. 
  • The process of planning and developing a new filter.
  • How bias manifests in image-trained models, and how Kit and his team are mitigating this.  
  • The most interesting and game-changing applications that Kit has worked on. 
  • His advice to other leaders of AI-powered startups.
  • Kit’s vision for the future of Plainsight Technologies.


Quotes:

“Our goal is to give customers very high accuracy on their models.” — Kit Merker


“A lot of times, traditional enterprises are looking for a solution or an app. The filter is like an app, and so customers can start really small with us, get an app that they trust the data, and then expand from there. They don't have any machine learning expertise required.” — Kit Merker


“Don't fake your demos!” — Kit Merker


Links:

Kit Merker

Kit Merker on LinkedIn

Kit Merker on X  

Plainsight Technologies


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Interpreting Infant Cries with Charles Onu from Ubenwa Health15 Jul 202400:22:53

Infants cry when they're hungry, tired, uncomfortable, or upset. They also cry when they’re in pain or severely ill. But how can parents tell the difference? To help us address this critical question, I'm joined by Charles Onu, a health informatics researcher, software engineer, and CEO of Ubenwa. Ubenwa is a groundbreaking app that uses AI to interpret infants' needs and health by analyzing the biomarkers in their cries. Charles conceived of the idea while working in local communities in south-eastern Nigeria, where high rates of newborn mortality due to late detection of Perinatal Asphyxia inspired him to create a solution.

In this episode, Charles shares insights into Ubenwa's machine-learning models and how they establish an infant's cry as a vital sign. He discusses the process of collecting and annotating data through partnerships with children's hospitals, the challenges of working with audio data, the benefits of creating a foundation model for infant cries, and much more. He also offers human-focused advice for leaders of AI-powered startups and reflects on his vision for success and the impact he hopes to achieve with Ubenwa. Tune in to discover how understanding your infant’s cries can transform healthcare and well-being for newborns and their families!


Key Points:

  • Charles' converging interests in math and healthcare, which led him to create Ubenwa.
  • What Ubenwa does to establish an infant’s cry as a vital sign (and why it’s so important).
  • The essential end-to-end role that machine learning plays in this technology.
  • How Ubenwa collects and annotates data by partnering with children’s hospitals.
  • Challenges of working with audio data and training medical ML models on it.
  • Insight into the benefits of creating a foundation model for infant cries.
  • Variations in infant’s cries and how Ubenwa’s models generalize for these shifts.
  • Valuable research Ubenwa has made publicly available as a gift to the ML community.
  • Charles’ human-focused advice for other leaders of AI-powered startups.
  • What success means to Charles and the impact he hopes to make with Ubenwa.


Quotes:

“Ubenwa was born out of the idea that, if there's something that [human doctors] can listen to to come to a conclusion [about an infant’s health], then there has to be something machines can also learn from the infant's cry.” — Charles Onu


“The real leap we made with self-supervised learning is that you now do not need an external annotation to learn. The model can use the data to supervise itself.” — Charles Onu


“AI-powered or not, – the problem of a startup remains the same. It’s to meet a need that humans have. – At the end of the day, AI is not just there for AI only. It’s only going to be a successful and useful startup if you identify a need and [solve] that problem.” — Charles Onu


“Human babies have evolved to communicate their needs and their health through their cries. We [haven’t] had the tools to understand that. Babies have been trying to talk to us for a long time. It's time to listen.” — Charles Onu


Links:

Ubenwa Health

Nanni AI

Charles Onu on LinkedIn

Charles Onu on X

Charles Onu on GitHub

Ubenwa on GitHub

Ubenwa CryCeleb Database


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Remote Monitoring and Water Forecasting with Marshall Moutenot from Upstream Tech08 Jul 202400:26:35

Innovative AI technologies are paving the way for more efficient and impactful environmental monitoring. Joining me today to discuss remote monitoring and water forecasting is Marshall Moutenot, the co-founder and CEO of Upstream Tech. From using satellite imagery to monitor conservation projects to employing machine learning for accurate water flow predictions, Upstream Tech is at the forefront of leveraging technology to address environmental challenges.

In our conversation, Marshall shares his journey from a tech-savvy childhood to co-founding a company with a mission to make environmental monitoring scalable and cost-effective. He delves into the development of Upstream Tech's two primary products: Lens, for remote monitoring of climate solutions, and HydroForecast, which uses AI to predict water flow, aiding in hydropower management. Marshall also underscores the need for integrating domain knowledge with machine learning to create reliable models before offering practical insights for AI startups. Tune in to learn more about how AI can revolutionize environmental conservation!


Key Points:

  • The details of Marshall’s tech-savvy childhood and entrepreneurial journey.
  • An overview of Upstream Tech’s mission to improve environmental monitoring.
  • How they use AI and satellite imagery for scalable, cost-effective monitoring.
  • The development of their Lens product for remote monitoring of climate solutions.
  • Why remote monitoring is so challenging at scale and their approach to solving it.
  • Their product, HydroForecast, and its role in predicting water flow using machine learning.
  • How integrating new inputs like satellite imagery creates reliable, adaptable models.
  • Success stories, including outperforming traditional models in a major competition.
  • Challenges Upstream Tech faces in acquiring and integrating geospatial data.
  • Best practices for ensuring model reliability and effectiveness over time.
  • Their team's approach to developing a new machine learning product or feature.
  • Marshall’s advice for AI startups: don’t get too attached to the tools!
  • His vision for Upstream Tech’s impact on environmental conservation.


Quotes:
"What these new machine learning models that we're employing allow us to do is to provide enough data to the model to create [equations] to describe physical interactions." — Marshall Moutenot


“[The] adaptability of these models is something that is really exciting for the field overall." — Marshall Moutenot


"We train a single model on a wide diversity, which forces the model to learn the common rules across all of them.” — Marshall Moutenot

“As an organization, one of [Upstream Tech’s] purposes is to see the 100% renewable grid become a reality. We want to continue to contribute to that and to build forecasts that enable that future.” — Marshall Moutenot


Links:

Marshall Moutenot on LinkedIn

Marshall’s Blog

Upstream Tech

Upstream Tech on LinkedIn

Upstream Tech on X

Upstream Tech on YouTube


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Scaling Healthcare Through Virtual Primary Care with Anitha Kannan from Curai01 Jul 202400:22:40

What will it take to bring affordable, accessible, and timely healthcare to all? Curai, an AI-powered virtual clinic, is on a mission to do just that by leveraging AI to enhance the efficiency of licensed physicians through text-based virtual primary care. In today’s episode, I sit down with Anitha Kannan, head of AI and founding member of Curai, to talk about the transformative potential of virtual primary care and its role in scaling healthcare access.

In our conversation, Anitha delves into the technical aspects of using large language models for patient data processing, the challenges of training models with clinical data, and the strategies Curai employs to ensure high-quality care. We also discuss the innovative ways Curai integrates AI into healthcare, the significance of multidisciplinary teams, and Anitha’s vision for the future of virtual care. Tune in for an insightful conversation on scaling healthcare through virtual primary care and learn how Curai is making a real impact!

Key Points:

  • Some background on Anitha Kannan, and how she joined Curai.
  • An overview of Curai’s services as a virtual healthcare practice.
  • How they provide affordable and timely healthcare access through AI-enhanced systems.
  • Machine learning’s role in history taking, information gathering, and summarization.
  • How AI streamlines the workflow for physicians.
  • Their use of large language models to process patient data.
  • Training model challenges: ensuring clinical correctness and handling data omission issues.
  • Best practices they’ve developed for validating models and the importance of evaluation.
  • Fundamental differences between their work and how other LLMs, like ChatGPT, are trained.
  • Their strategy for balancing long-term research aspirations with short-term product development.
  • An overview of their multidisciplinary teams and how this contributes to their success.
  • Anitha’s hopes for the future of Curai; particularly through partnerships with healthcare organizations.


Quotes:

"Our mission is to provide the best health care to everyone." — Anitha Kannan


“Today, [Carai runs] a text-based virtual primary care practice. We have our licensed physicians or experts in their fields. Then we supercharge them and bring about a lot of efficiencies by leveraging AI.” — Anitha Kannan


"It's very easy to build 80% of a good product with AI today, but I think to get it to 100%, [and] to get it to scale, to be useful in [the] real world — evaluation is the number one thing." — Anitha Kannan


“At Curai, the AI team is composed of clinical experts, subject matter experts, researchers, and machine learning engineers. Every project, long-term or short-term, has a mix of these types of expertise in it. This allows us to work through the problem much more effectively.” — Anitha Kannan


Links:

Anitha Kannan on LinkedIn 

Anitha Kannan on X

Curai Health


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Better EV Batteries with Jason Koeller from Chemix24 Jun 202400:27:30

Batteries are arguably the most important technological innovation of the century, powering everything from mobile phones to electric vehicles (EVs). Unfortunately, most batteries have a significant impact on the environment, requiring increasingly scarce and valuable resources to manufacture and typically not designed for easy repair, reuse, or recycling.

Today on Impact AI, I'm joined by Jason Koeller, Co-Founder and CTO of Chemix, to find out how his company is leveraging AI to create better, more sustainable EV batteries that could reduce our reliance on elements like lithium, nickel, and cobalt, all without compromising vehicle performance. For a fascinating conversation with a data-driven physicist working at the intersection of software, machine learning, chemistry, and materials science, be sure to tune in today!


Key Points:

  • Jason’s background in theoretical physics and how it led him to create Chemix.
  • Products and services offered by Chemix and the role that AI plays.
  • Four reasons that machine learning (ML) is at the core of everything Chemix does.
  • Unique challenges that their ML models need to contend with.
  • What goes into validating these models to ensure accuracy.
  • Why now is the right time for the technology that Chemix is developing.
  • Metrics for measuring the impact of a better EV battery.
  • Jason’s data-driven advice for leaders of AI-powered startups.
  • His “electrifying” vision for Chemix in the next three to five years.


Quotes:

“All data analysis and decision-making is automated by our AI system. This includes analyzing terabytes of battery test data each day.” — Jason Koeller


“Looking at broad trends, [electric vehicles (EVs)] and AI have both become [things] that people have been talking a lot more about in the past 10 years and even more so in the past four or five years, and that has happened simultaneously.” — Jason Koeller


“Why is everyone not buying an EV? It's largely because they're too expensive or because people are worried they're not charging fast enough or they don't hold enough range for long road trips. – Improving any one of these metrics would be a measure of impact.” — Jason Koeller


Links:

Jason Koeller on LinkedIn

Chemix

Chemix on LinkedIn


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Foundation Model Series: Accelerating Pathology Model Development Using Embeddings with Julianna Ianni from Proscia27 Jan 202500:20:51

How can foundation models accelerate breakthroughs in precision medicine? In today’s episode of Impact AI, we explore this question with returning guest, Julianna Ianni, Vice President of AI Research and Development at Proscia, a company revolutionizing pathology through cutting-edge technology. Join us as we explore how their platform, Concentriq, and its new Embeddings feature are transforming AI model development, making pathology-driven insights faster and more accessible than ever before. You’ll also learn how Proscia is shaping the future of precision medicine and discover practical insights for leveraging AI to advance healthcare. Whether you're curious about pathology, AI, or innovations in precision medicine, this episode offers invaluable takeaways you won’t want to miss!


Key Points:

  • An overview of Julianna’s biomedical engineering background and Proscia's mission.
  • Insight into Proscia’s Concentriq platform, aiding more than two million diagnoses annually.
  • Ways that Concentriq Embeddings streamlines AI development by eliminating data friction.
  • How Concentriq Embeddings make model creation 13x faster than traditional methods.
  • Why Proscia integrates external foundation models for versatility and superior performance.
  • Flexible and efficient: how Concentriq lets users test, swap, and select models with ease.
  • Types of solutions built using these embeddings, including rapid biomarker detection.
  • Tackling AI challenges like reducing overfitting and addressing bias in medical applications.
  • Lessons from pathology: simplifying complex workflows for faster AI adoption in other fields.
  • A look at the future of foundation models for pathology and Julianna’s advice for innovators.


Quotes:

“With the rise of foundation models that are pathology-specific and more powerful than the models of yesterday, the ability to extract embeddings efficiently became even more important for us.” — Julianna Ianni


“The pathology world didn't need another hit movie. It needed a streaming service.” — Julianna Ianni


“[Continue] to innovate and [understand] what's out there. There's a lot of change in the [pathology] field right now – You're going to make plans and then you're going to need to remake those plans because things are changing so quickly.” — Julianna Ianni


“ChatGPT didn't pervade our culture because it's fantastic technology. It pervaded our culture because the fantastic technology was easy to use. Pathology should be that easy. Our aim is to drive it there.” — Julianna Ianni


Links:

Proscia

Julianna Ianni on LinkedIn

Julianna Ianni on X

Julianna Ianni on Google Scholar

Concentriq Embeddings
Concentriq Embeddings internal case study
Proscia AI Toolkit
Zero-Shot Tumor Detection Example

Previous episode of Impact AI: Data-Driven Pathology with Coleman Stavish and Julianna Ianni from Proscia


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Personalized Cancer Treatment Decisions with Nathan Silberman from Artera17 Jun 202400:17:10

Being given a cancer diagnosis is one of the worst pieces of news you can receive as a patient. This is often made even more difficult by the fact that choosing a treatment option is rarely simple or easy. Clinicians need to make multiple assessments before they can move forward, and even then it is often difficult or impossible to make unambiguous predictions. That’s where Artera comes in, a company using multimodal AI tests to provide individualized results for cancer patients, which enables clinicians and patients to make personalized treatment decisions, together.

I am joined today by Nathan Silberman, Vice President of Machine Learning and Engineering at Artera, to talk about how Artera’s technology is paving the way for personalized cancer treatment decisions. Join us today, as we get into how Artera is contributing to the cancer treatment process, some of the biggest challenges they face, and how they are addressing these through specifically trained algorithms and robust validation protocols. Be sure to tune in to this important conversation on how Artera is impacting cancer treatment outcomes for the better!


Key Points:

  • Background on our guest, Nathan Silberman, and what led him to Artera.
  • How Artera is helping clinicians make informed decisions for cancer treatments.
  • The role of machine learning in their personalized risk assessments for patients.
  • Key challenges they’ve encountered with pathology data.
  • How they deal with slide variations through well-trained algorithms.
  • Bias in pathology data and what Artera is doing to mitigate bias.
  • Their partnerships with academics, clinicians, and oncologists.
  • Insight into the variety of approaches they use to validate their models.
  • How their tests fit in with clinical workflows and assist doctors and patients.
  • The agonizing wait time associated with traditional non-AI testing methods.
  • How Artera is providing quick and reliable test results.
  • Advice to leaders of AI-powered startups: stay focused on the ultimate goal of patient impact.
  • Looking ahead at Artera’s impact in the next three to five years.


Quotes:

“Which therapy to choose is simply not an easy choice. Clinicians would ideally be able to accurately assess a patient's risk of a cancer spreading, or adversely affecting the patient's health in the short term. But often, that's hard or impossible for a clinician to predict.” — Nathan Silberman


“Clinicians have been wanting and waiting for tools that can predict whether or not a therapy will work for that particular patient. This is ultimately where Artera steps in.” — Nathan Silberman


“Rather than wait a month, Artera's test provides the answer within two to three days after the lab receives the biopsy slide. And it is so rewarding to hear from clinicians, and especially patients about the relief we can provide by giving clarity sooner.” — Nathan Silberman


“I think the biggest piece of advice I can give is really just making sure that you're laser-focused on the ultimate goal of patient impact.” — Nathan Silberman


Links:

Artera

Nathan Silberman on LinkedIn


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Foundation Model Assessment – Foundation models are popping up everywhere – do you need one for your proprietary image dataset? Get a clear perspective on whether you can benefit from a domain-specific foundation model.

Faster Object Search with Corey Jaskolski from Synthetaic10 Jun 202400:27:12

What if there was a way to revolutionize image-based AI, eliminating the need for extensive prework? In this episode, I sit down with Corey Jaskolski, Founder and President of Synthetaic, to talk about finding objects in images and video quickly. Synthetaic is redefining the landscape of data analysis with its groundbreaking technology that eliminates the need for time-consuming human labeling or pre-built models. It specializes in the rapid analysis of large, unlabeled video and image datasets.

In our conversation, we delve into the groundbreaking technology behind Synthetaic's flagship product and how it is revolutionizing image and video processing. Explore how it utilizes an unsupervised backend to swiftly analyze and interpret data, how it is able to work with any kind of image data, and the process behind ingesting and embedding image objects. Discover how Synthetaic navigates biased data and leverages domain expertise to ensure accurate and ethical AI solutions. Gain insights into the gaps holding AI’s application to images back, the different ways the company’s technology can be applied, the future development of Synthetaic, and more!


Key Points:

  • Corey’s background in AI and ML and what led to the creation of Synthetaic.
  • Why Synthetaic focuses on processing images and videos quickly.
  • How the company leverages ML in its approach. 
  • Details about image ingestion and embedding processes.
  • How the definition of potential objects varies depending on the type of imagery used.
  • Explore the role of domain expertise in addressing challenges. 
  • Hear examples of the technology’s diverse range of applications.
  • Recommendations to leaders of AI-powered startups. 
  • His hope for the future trajectory of Synthetaic.


Quotes:

“We think about the machine learning problems a little bit differently, because we're not labeling data to go ahead and build a bespoke frozen traditional AI model.” — Corey Jaskolski


“We take this very broad view of objects where anything that could be discrete from anything else in the imagery gets called an object, at the risk of basically finding, if you will, too many objects.” — Corey Jaskolski


“We think of RAIC as something that solves the cold start problem really well.” — Corey Jaskolski


“By and large, we're training image and video-based AIs the same way. We need a paradigm shift that really allows AI to be the force multiplier that it can be.” — Corey Jaskolski


Links:

Corey Jaskolski on LinkedIn

Corey Jaskolski on X

Synthetaic


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Foundation Model Assessment – Foundation models are popping up everywhere – do you need one for your proprietary image dataset? Get a clear perspective on whether you can benefit from a domain-specific foundation model.

Digital Twins for Clinical Trials with Charles Fisher from Unlearn AI03 Jun 202400:30:19

What if AI could improve the outcomes of clinical trials by making them more efficient and reducing the number of patients receiving placebos? Well, today’s guest, Charles Fisher is here to tell us all about how his company, Unlearn AI, is creating digital twins to do just that! In this conversation, you’ll hear all about Charles' academic background, what made him decide to create Unlearn AI, what the company does, and how they work within clinical trials. We delve into the problems they focus on and the data they collect before Charles tells us about their zero-trust solution. We even discuss Charles’ opinions of how domain knowledge should be used in machine learning. Finally, our guest shares advice for leaders of AI-powered startups. To hear all this and even find out what to expect from Unlearn in the near future, tune in now!


Key Points:

  • A rundown of Charles Fisher’s background and what led him to create Unlearn AI. 
  • What Unlearn does, what digital twins are, and why they’re important. 
  • How clinical trials work and how they are used within Unlearn. 
  • The kinds of data they use and how they tackle these clinical trials using machine learning. 
  • What a zero-trust solution is and how Unlearn guarantees that their results are accurate. 
  • Charles shares his thoughts on the role of domain expertise in machine learning. 
  • His advice for any leaders of AI-powered startups. 
  • What we can expect from Unlearn in the next three to five years. 


Quotes:

“[Unlearn is] typically working on running clinical trials where we might be able to reduce the number of patients who get the placebo by somewhere like – 50%.” — Charles Fisher


“[Unlearn] can prove that these studies produce the right answer, even though they leverage these AI algorithms.” — Charles Fisher


“It's very difficult to find examples where you can actually have a zero-trust application of AI. I actually don't know of another one besides [Unlearn’s].” — Charles Fisher


Links:

Charles Fisher on LinkedIn

Charles Fisher on X

Unlearn AI


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Foundation Model Assessment – Foundation models are popping up everywhere – do you need one for your proprietary image dataset? Get a clear perspective on whether you can benefit from a domain-specific foundation model.

Cutting Carbon in Concrete with Mathieu Bauchy from Concrete.ai27 May 202400:30:42

Did you know that concrete is the second most-used material in the world after water? Although it has largely defined modern society, concrete has a hidden climate cost: it is responsible for 1.6 billion tons of carbon dioxide entering the atmosphere annually. For context, that’s more than the entire aviation industry! With these statistics in mind, today’s guest is on a mission to decarbonize the construction industry. As the CTO and co-founder of cleantech startup, Concrete.ai, Mathieu Bauchy is using his expertise in artificial intelligence and materials modeling to prescribe new concrete formulations that are less carbon-intensive and more economical. Today, Mathieu joins me to offer insight into Concrete.ai's exciting technology, why it’s important for the planet, and how it can reduce concrete emissions by a third while also ensuring that concrete producers maximize margins and streamline their supply chains. To find out how this is possible without any changes to the raw materials, no modification of the production process, and no cost premium, be sure to tune in today!


Key Points:

  • Insight into Mathieu’s research focus and how it led him to create Concrete.ai.
  • What Concrete.ai does and why it’s important for reducing CO2 emissions.
  • The role of machine learning, particularly generative AI, in this technology.
  • How Concrete.ai develops ML models that are reliably able to extrapolate.
  • Why estimating uncertainty is important and how Concrete.ai approaches it.
  • What goes into validating these models, including systematic testing in the field.
  • Reasons that the timing for Concrete.ai’s technology is critical.
  • Dollars saved and other metrics for measuring the impact of this technology.
  • Mathieu’s humanity-focused advice for other leaders of AI-powered startups.
  • How Concrete.ai’s impact will continue to expand and evolve.


Quotes:

“Concrete is responsible for 8% of the total CO2 emissions in the world. To give you some context, that's about three times more emissions than the entire aviation industry.” — Mathieu Bauchy


“We think that it's the right time for the concrete industry to benefit from what AI can offer to avoid waste during the production of concrete. The idea is that, if we adopt these new technologies, then we can continue to improve our quality of life.” — Mathieu Bauchy


“It's not like we are changing the way concrete is made. It's still made in the same plant. It's still made using the same materials. We are just changing the recipe, and just that [can] save about a third of the emissions of concrete.” — Mathieu Bauchy


“AI also comes with its own carbon footprint and, to some extent, also contributes to climate change. We should think about how we use AI to solve climate change and not further contribute to it.” — Mathieu Bauchy


Links:

Concrete.ai
Concrete.ai on LinkedIn

Mathieu Bauchy

Mathieu Bauchy on LinkedIn

Mathieu Bauchy on YouTube

Mathieu Bauchy on X


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Foundation Model Assessment – Foundation models are popping up everywhere – do you need one for your proprietary image dataset? Get a clear perspective on whether you can benefit from a domain-specific foundation model.

Decoding Pathology for Precision Medicine with Maximilian Alber from Aignostics20 May 202400:19:35

Today, I am joined by Maximilian Alber, Co-founder and CTO of Aignostics, to talk about pathology for precision medicine. You’ll learn about Aignostics’s mission, how they are impacting healthcare, and the transformative power of foundational models. Max explains how Aignostics is driven by the belief that machine learning and data science will help improve healthcare before expanding on the role of foundational models. He describes how they built their foundational model, what sets it apart from other models, and why diversity in their datasets is key. He also breaks down how foundational models have allowed them to develop other models more quickly and better navigate explainability with concepts that are challenging for machine learning. We wrap up with Max’s advice for leaders of other AI-powered startups and where he expects Aignostics will be in the next five years. Tune in now to learn all about foundational models and the innovative work being done at Aignostics!


Key Points:

  • Insight into Max’s role at Aignostics and how the company is impacting healthcare.
  • How they use machine learning to set themselves apart from their competitors.
  • A rundown of their models and datasets.
  • The definition of a foundation model and how Aignostics built theirs.
  • How to use foundation models as a starting point for building machine learning applications.
  • What sets Aignostics’ foundation model for histopathology apart from other similar models.
  • How their foundation model enables them to develop other models more quickly.
  • Top lessons Max has learned from developing foundation models.
  • How they navigate explainability with concepts that are challenging for machine learning.
  • The positive impact that foundational models have had on explainability.
  • Recent advancements that Max is excited about as potential use cases for Aignostics.
  • Max’s advice to leaders of other AI-powered startups.
  • The impact of Aignostics and where he expects it will be in the next three to five years.


Quotes:

“Our mission is to turn biomedical data into insights.” — Maximilian Alber


“Everything we do is driven by the belief that machine learning and data science will help us improve healthcare.” — Maximilian Alber


“A foundation model is a model that can be used as a starting point for building a machine learning application, with the promise that the foundation model already has a great understanding of the domain.” — Maximilian Alber


“We are in active discussions for licensing our foundation model to other companies in order to enable their development as well. [What’s] important here is that we develop our foundation model along regulatory requirements, which will allow it to be used in medical products.” — Maximilian Alber


“One needs to build a technology that either makes a difference in the long run, or one must be able to innovate at a very fast pace.” — Maximilian Alber


Links:

Maximilian Alber on LinkedIn

Aignostics

Aignostics on LinkedIn


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Foundation Model Assessment – Foundation models are popping up everywhere – do you need one for your proprietary image dataset? Get a clear perspective on whether you can benefit from a domain-specific foundation model.

Subseasonal-to-Seasonal Weather Forecasting with Sam Levang from Salient Predictions13 May 202400:16:51

Advanced weather forecasts are the new frontier in meteorology. Long-term forecasting has garnered significant attention due to its potential to provide valuable insights to various sectors of society and the economy. In today’s episode, Sam Levang, Chief Scientist at Salient, joins me to discuss Salient’s innovative approach to weather forecasting. Salient specializes in providing highly accurate subseasonal-to-seasonal weather forecasts ranging from 2 to 52 weeks in advance.

In our conversation, we discuss the ins and outs of the company’s innovative approach to weather forecasting. We delve into the hurdles of subseasonal-to-seasonal forecasting, how machine learning is replacing traditional weather modeling approaches, and the various inputs it uses. Discover the value of machine learning for post-processing of data, the type of data the company utilizes, and why it uses probabilistic models in its approach. Gain insights into how Salient is catering to the impacts of climate change in its weather predictions, the company’s approach to validation, how AI has made it all possible, and much more!


Key Points:

  • Sam's background in science and the creation of Salient.
  • Hear how Salient is revolutionizing weather forecasting and why.
  • How Salient is utilizing machine learning in its forecasting models.
  • Examples of the data and models the company uses.
  • The challenges of working with weather data to build models.
  • Explore why Salient also uses probabilistic models in its approach.
  • Salient’s approach to validation and how it deals with data uncertainty.
  • Ways AI has made the company’s approach to forecasting possible. 
  • He shares advice for leaders of other AI-powered startups.


Quotes:

“Salient produces weather forecasts that extend further into the future than most people are used to seeing. We go up to a year in advance.” — Sam Levang


“ML (Machine Learning) models have proved to be actually a very effective replacement for the traditional approach to weather modeling.” — Sam Levang


“The only difference about making forecasts longer timescales of weeks and months ahead is that there are some differences in the particular parts of the climate system that provide the most predictability.” — Sam Levang


“While ML and AI are extremely powerful tools, they are still just tools and there's so much else that goes into building a really valuable product, or a service, or a company.” — Sam Levang


Links:

Sam Levang on LinkedIn 

Salient

Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Foundation Model Assessment – Foundation models are popping up everywhere – do you need one for your proprietary image dataset? Get a clear perspective on whether you can benefit from a domain-specific foundation model.

Virtual Tissue Staining with Yair Rivenson from PictorLabs06 May 202400:34:11

Welcome to today’s episode of Impact AI, where we dive into the groundbreaking world of virtual tissue staining with Yair Rivenson, the co-founder and CEO of PictorLabs, a digital pathology company advancing AI-powered virtual staining technology to revolutionize histopathology and accelerate clinical research to improve patient outcomes. You’ll find out how machine learning is used to translate unstained tissue autofluorescence into diagnostic-ready images, gain insight into overcoming AI hallucinations and the rigorous validation processes behind virtual staining models, and discover how PictorLabs navigates challenges like large files and bandwidth dependency while seamlessly integrating technology into clinical workflows. Yair also provides invaluable advice for AI-powered startup leaders, emphasizing the importance of automation and data quality. To gain deeper insights into the transformative potential of virtual tissue staining, tune in today!


Key Points:

  • The origin story of PictorLabs and the research that informed it.
  • Why Pictor’s work is so important for patients and the healthcare system.
  • What Yair means when he says machine learning is the “engine” for virtual staining.
  • How Pictor mitigates the challenge of AI hallucinations.
  • Insight into what goes into validating virtual staining models.
  • Large files, bandwidth dependency, and other challenges that Pictor faces.
  • A look at how this technology fits smoothly into the clinical workflow.
  • Collaborating with economic partners while staying focused on business objectives.
  • Yair’s product-focused advice for leaders of AI-powered startups
  • What the next three to five years looks like for PictorLabs.


Quotes:


“The most important factor for the healthcare system, for the patient is the fact that you can get all the results, all the workup, and all the different stains from a single tissue section very, very fast.” — Yair Rivenson


“Machine learning is the engine behind virtual staining. In a sense, that’s what takes those images from the autofluorescence of the unstained tissue section and converts [them] into a stain that pathologists can use for their diagnostics.” — Yair Rivenson


“At the end of the day, the network is as good as the data that it learns from.” — Yair Rivenson


“The more you automate, the better off you’ll be in the long run.” — Yair Rivenson


Links:

Yair Rivenson

PictorLabs

PictorLabs on LinkedIn

‘Virtual histological staining of unlabelled tissue-autofluorescence images via deep learning’

‘Assessment of AI Computational H&E Staining Versus Chemical H&E Staining For Primary Diagnosis in Lymphomas’


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Foundation Model Assessment – Foundation models are popping up everywhere – do you need one for your proprietary image dataset? Get a clear perspective on whether you can benefit from a domain-specific foundation model.

Improving Recycling Efficiency with Nikola Sivacki from Greyparrot29 Apr 202400:20:31

One of the most powerful impacts machine learning can make is helping to solve environmental challenges all around the world. Today on Impact AI, I am joined by the founder of Greyparrot, Nikola Sivacki to discuss how his company uses machine learning to improve recycling efficiency. Learn all about Nikola’s background, what Greyparrot does, their services, the importance of their work, the role machine learning plays in it, how they gather and annotate data, the challenges they face, how they develop new models, and so much more. Tune in to hear the newest AI innovations Nikola is most excited about before hearing his goals for Greyparrot in the near future. Lastly, get some valuable advice for running AI-powered startups.


Key Points:

  • Welcoming Nikola Sivacki to the show. 
  • Nikola shares a bit about his background and how it led him to create Greyparrot. 
  • What Greyparrot does, what services they offer, and why it is so important. 
  • The role machine learning plays in this technology. 
  • How they go about gathering data and annotating it for their purposes. 
  • What they are trying to predict with the data they are gathering. 
  • Challenges they encounter in training machine learning models and how to overcome them.
  • A breakdown of how his team plans and develops a new machine learning model or feature. 
  • Nikola shares how Greyparrot measures the impact of its technology. 
  • The two groups of machine learning developments Nikola is most excited about. 
  • Nikola shares some advice for other leaders of AI-powered startups. 
  • Where he sees the impact of Greyparrot in three to five years. 


Quotes:

“Greyparrot basically monitors the flow of waste materials, recyclable materials in material recovery facilities, and offers compositional analysis of these materials.” — Nikola Sivacki


“It's very helpful, – if thinking of a new product, to start with a data set that is really tailored to answering the main uncertain question that is posed there.” — Nikola Sivacki


“Start thinking about data from the start. I think that it’s very important to understand the data in detail.” — Nikola Sivacki

“Our goal is to improve, of course, recycling rates globally so that we can reduce reliance on virgin materials.” — Nikola Sivacki


Links:

Nikola Sivacki on LinkedIn

Nikola Sivacki on X

Greyparrot


Resources for Computer Vision Teams:

LinkedIn – Connect with Heather.

Computer Vision Insights Newsletter – A biweekly newsletter to help bring the latest machine learning and computer vision research to applications in people and planetary health.

Computer Vision Strategy Session – Not sure how to advance your computer vision project? Get unstuck with a clear set of next steps. Schedule a 1 hour strategy session now to advance your project.

Foundation Model Assessment – Foundation models are popping up everywhere – do you need one for your proprietary image dataset? Get a clear perspective on whether you can benefit from a domain-specific foundation model.

© My Podcast Data