Data Science Conversations – Details, episodes & analysis

In this episode, we had the privilege of speaking with Walid Mehanna, Chief Data and AI Officer at Merck Group. Walid shares deep insights into how large, complex organizations can scale data and AI and create lasting impact through thoughtful leadership.

As Chief Data & AI Officer of Merck Group, Walid led the Merck Data & AI Organization, delivering strategy, value, architecture, governance, engineering, and operations across the whole company globally. Hand in hand with Merck’s business sectors and their data offices, we harnessed the power of Data & AI. Walid is glad to be part of Merck as another curious mind dedicated to human progress.

In our latest episode of the Data Science Conversations Podcast, we spoke with Christoph Sporleder, Managing Partner at Rewire, about the evolving role of consulting in the data and AI space.

This conversation is a must listen for anyone dealing with the challenges of integrating AI into business processes or considering an AI project with an external consulting firm. Christoph draws from decades of experience, offering practical advice and actionable insights for organizations and practitioners alike.

Key Topics Discussed

1. Evolution of Data and Cloud Computing

The shift from local computing to cloud technologies, enabling broader data integration and advanced analytics, with the rise of IoT and machine data.

2. Data Management Challenges

Discussion on the evolution from data warehouses to data lakes and the emerging concept of data mesh for better governance and scalability.

3. Importance of Strategy in AI

Why a clear strategy is crucial for AI adoption, including aligning organizational leadership and identifying impactful use cases.

4. Sectoral Adoption of Data and AI

Differences in adoption across sectors, with early adopters in finance and insurance versus later adoption in manufacturing and infrastructure.

5. Consulting Models and Engagement

Insights into consulting engagement types, including strategy consulting, system integration, and body leasing, and their respective challenges and benefits.

6. Challenges in AI Implementation

Common pitfalls in AI projects, such as misalignment with business goals, inadequate infrastructure planning, and siloed lighthouse initiatives.

7. Leadership’s Role in AI Success

The critical need for senior leadership commitment to drive AI adoption, ensure process integration, and manage organizational change.

8. Effective Collaboration with Consultants

Best practices for successful partnerships with consultants, including aligning on objectives, managing personnel transitions, and setting clear engagement expectations.

9. Future Trends in Data and AI

Emerging trends like componentized AI architectures, Gen AI integration, and the growing focus on embedding AI within business processes.

10. Tips for Managing Long-Term Projects

Strategies for handling staff rotations and maintaining project continuity in consulting engagements, emphasizing planning and communication.

In this episode Patrick McQuillan shares his innovative Biological Model - a concept you can use to enhance data outcome in large enterprises.  The concept takes the idea that the best way to design a data strategy is to align it closely with a biological system.

He discusses the power of centralized information, importance of data governance, and the necessity for a common performance narrative across an organization.

Episode Summary -

- Biological Model Concept

- Centralized vs. Decentralized Data

- Data Collection and Maturity

- Horizontal translation layer 

- Partnership with vertical leaders

 - Curated data layers 

- Data dictionary for consistency

- Focusing on vital metrics

- Data Flow in Organizations

- Biological Model Governance

- Overcoming Inconsistency and Inaccuracy

In this episode Heidi Hurst returns to talk to us about how in her current role at Pachama she is using the power of machine learning to fight climate change.  She discusses her work in measuring the capacity of existing forests and reforestation projects using satellite imagery.

Episode Summary

1. The importance of carbon credits verification in mitigating climate change

2. How Pachama is using machine learning and satellite imagery to verify carbon projects

3. Three types of carbon projects: avoided deforestation, reforestation, and improved forest management

4. Challenges in using satellite imagery to measure the capacity of existing forests

5. The role of multispectral imaging in measuring density of forests

6. Challenges in collecting data from dense rainforests and weather obstructions

7. The impact of machine learning on scaling up carbon verification

8. Advancements in the field of satellite imaging, particularly in small satellite constellations

Ágnes Horvát is an Assistant Professor in Communication and Computer Science at Northwestern University. Her work focuses on understanding how online networks induce biased information production, sharing and processing across digital platforms. 

- The new Post-normal era for science - Having an awareness of the context and values that impact scientific research

• Where is science communication in relation to digital platforms? - Scholars work hard on discovering scientific findings, but information doesn’t always reach the public appropriately.  

• How to communicate scientific research - it’s not just about communicating with scientists and general audiences. News needs to reach policymakers and governments too for real change.

• The production of scientific research has exploded recently thanks to decision-making demands - and the pandemic had a lot to do with this. Scientists were under pressure to carry out research quickly and at the expense of quality. 

• Misinformation can have detrimental consequences - even leading to vaccine hesitancy in some communities.

• The surprising effect of retracting papers - papers that get retracted in the future are more likely to receive more engagement before getting withdrawn.

• Why are paper retractions on the rise?  - again, the recent pandemic has caused an increase in retractions. 

• Is social media helping or hindering science research? - while the platforms are helping to  spread real news, social media also helps the spread of false information. 

• As long as you have quality data and robust trends - regardless of the method, you will identity that trend. 

Reducing the problem of miscommunication - with whom does the responsibility lie?

Modern Data Infrastructures and platforms store huge amounts of multidimensional data.  But - data pipelines frequently break and a machine learning algorithm's performance is only as good as the quality and reliability of the data itself.

In this episode we are joined by Lior Gavish and Ryan Kearns of Monte Carlo, to talk about how the new concept of Data Observability is advancing Data Reliability and Data Quality at Scale.

Episode Summary

1. A overview of Data Reliability/Quality and why it is so critical for organisations
2. The limitations of traditional approaches in the area of Data Reliability
3. Data observability and why it is different to traditional approaches to Data Quality
4. The 5 Pillars of Data Observability
5. How to improve data reliability/quality at scale and generate trust in data with stakeholders.
6. How observability can lead to better outcomes for Data Science and engineering teams?
7. Examples of data observability use cases in industry
8. Overview of O’Reilly’s upcoming book, The Fundamentals of Data Quality.

In this episode we are joined by Julia Stoyanovich from NYU, to talk about her work into how AI is being used in the hiring process.

Whether you are responsible for hiring on behalf of a business or are a job seeker, you will find this podcast very interesting, but for very different reasons.

Episode Summary

• Algorithmic decision making in the hiring process - what does that mean for businesses and job seekers?
• The hiring process - the funnel effect.
• Lack of public disclosure about the use of algorithmic tools as part of the talent acquisition pipeline.
• Are job seekers being unfairly screened out of the hiring process?
• How AI based implementations of psychometric instruments are used today.
• Is it possible to measure a person’s personality based on data alone?
• Do these systems remove bias and discrimination from the hiring process?
• Testing the stability and consistency of these algorithmic systems.
• Vendors of systems and their lack of testing / recognising the issues.
• Are new laws needed so the hiring process is fairer and more transparent?
• What does the future of hiring look like - fewer AI systems and more human intervention?

In this episode we are joined by Professor Maurizio Porfiri from NYU, to talk about his latest academic research which is using data science to uncover why sales of guns in the USA increase after a mass shooting event.  

His interest and research was borne from a very personal experience 14 years ago when he experienced a mass shooting event at Virginia Tech where he was studying.

• Researching Complex Systems
• Virginia Tech Mass Shooting event and its impact on Maurizio
• What is the relationship between mass shooting events and the purchase of guns?
• Analysis of time series data - 70 mass shootings in around 20 years
• Can media coverage on mass shootings shape public opinion, thereby influencing firearm acquisition?
• Examining the correlation between three distinct datasets
• What are the causes of increased gun sales in the aftermath of mass shooting events?
• Differences in the data at State level V National level?
• Researching the complex firearm ecosystem with all its pieces - prevalence, violence and regulation

In this episode we are joined by Perry Marshall to talk about his latest scientific paper entitled “Biology Transcends the Limits of Computation”.  We also discuss his $10 million Evolution 2.0 Science Prize, which is the largest prize in the world in science currently.

His paper pushes the boundaries in the field of evolutionary biology and his science prize is driving some truly fascinating and thought provoking implications for the development of strong AI.

In this episode we are joined by Arnon Houri Yafin, an Israeli entrepreneur who is the founder of a company called Zzapp Malaria, which recently won the AI XPRIZE sponsored by IBM Watson.

Their work in using AI to eliminate malaria in Africa is both interesting and inspirational.

Episode Summary

• Moving from malaria control to malaria elimination
• How Zzapp Malaria started and how investors were attracted
• The use of drones, satellite imagery, topography, rain/humidity data and a new mobile app
• The development of small neural networks to identify the potential for small water bodies
• How IBM Watson assisted with funding but also the machine learning models
• The use of biology agents rather than chemicals to treat stagnant water bodies
• The NGO project in Ghana - reduced mosquitos by 60% in 100 days
• Latest and biggest operation to date in São Tomé and Príncipe
• The direct impact on the GDP of a country’s economy as a result of eliminating malaria
• How malaria and poverty are interconnected
• Winning the $3 million XPrize and how the money will be used
• How you can help by supporting Malaria No More and Only Nets