Explore every episode of the podcast How to AI UXR

Dr. Llewyn Paine is an AI consultant, product strategist, and innovation leader with nearly two decades of experience working in emerging technology. As the principal of Llewyn Paine Consulting, she specialises in helping product leaders implement evidence-based rigour and responsible AI practices in their UX and design workflows. Her career includes leading initiatives on intelligent agents and physical AI at Microsoft and developing experimental media at Disney. Paine serves as the lead curator for Rosenfeld Media’s Designing with AI conference and frequently speaks at major institutions, including the Library of Congress.

How to AI UXR is supported by Strella, an AI-powered customer research platform that partners with you to build, moderate, and synthesise interviews, allowing you to go from question to actionable insights in just a few hours.

In This Conversation

As the How to AI UXR map shows, research teams are now using AI across every part of the research workflow and beyond. They’re drafting screeners, summarising interviews, shaping reports, coaching non-researchers, automatically routing research requests, building end-to-end research systems—and learning just how much those systems cost to maintain. Velocity is the name of the game, but the important question isn’t whether AI can make research faster; it’s whether research teams can move faster and preserve the judgement, evidence, and methodological care that make research valuable in the first place.

In this episode, Llewyn argues that AI in research should begin with judgement. Rather than treating AI adoption as a race towards more output, she makes the case for returning to the disciplines researchers already know well, including jobs to be done (JTBD), systems design, social science, statistics, and evaluation.

This episode is a useful corrective to the pressure many teams feel to “AI everything.” Llewyn’s view is not that researchers should reject AI or position themselves as blockers to progress; it’s that they should understand the tools well enough to assess risk, ask better questions, and decide where AI supports better outcomes rather than simply producing more material. This conversation will give you a clearer way to assess where AI belongs in your research system, what evidence to ask for, and what to watch for before you trust the work it helps produce.

The How to AI UXR Map

This series builds on the insights shared in the How to AI UXR map, a five-page map that charts key trends, helps you pinpoint your AI maturity level, and offers practical, real-world applications you can adapt to your research systems.

Download the Map

In this episode, we cover:

- Why a practice being popular does not make it good research practice

- The three lenses research teams need when evaluating AI: stakeholder needs, social science, and computer science

- Why teams should begin with the stakeholder’s job to be done before choosing an AI tool

- What an AI harness is, and why the model itself is only one part of the system (Llewyn shares a great Teenage Mutant Ninja Turtles analogy)

- Why Llewyn draws a distinction between AI that increases output and AI that improves outcomes

- Why qualitative synthesis is one of the most tempting and highest-risk uses of AI in research

- The difference between accuracy problems, such as incorrect quotes, and omission problems, where AI misses the most interesting or important insight

- Why operational use cases, including coaching, routing, training, and support, may be among the most valuable applications of AI for research teams

- Why researchers should understand enough about AI to ask for evidence, challenge weak claims, and avoid becoming passive consumers of vendor or influencer promises

- How token costs may push teams towards better system design, clearer context, and more thoughtful use of AI

- Why researchers do not need to carry evaluation work alone, and how they can partner with engineering and others to assess whether tools are doing what they claim

Partway through the episode, as a leader building one of the tools these researchers are exploring and using, Priya Krishnan, the cofounder and COO of Strella, shares her take on the conversation.

Things Referenced

- Jobs to be done (JTBD) is a framework for understanding what stakeholders are trying to accomplish, rather than beginning with a tool or deliverable.

- AI harnesses are the surrounding software infrastructure that gives a model access to tools, context, rules, workflows, memory, and guardrails.

- Evals (evaluations) are structured ways of testing whether an AI system is producing outputs that meet defined criteria.

- Markdown is a lightweight plain-text formatting language that helps humans and machines structure information clearly.

- CSV files are simple spreadsheet-style files often used to move structured data between tools.

- Tokens are the word fragments and other units of text that AI systems process as input and output.

- Designing with AI is a Rosenfeld Media conference that Llewyn helps curate.

- Paul Ford is a technology writer and software builder (best known for Bloomberg’s “What Is Code?”) who argues that AI makes human judgement and accountability more important, not less.

- Oen Michael Hammonds is a UX and AI practitioner who uses “AI speed bump” to describe adding deliberate friction and checks so teams don’t ship unsafe or unreliable AI systems.

- World Usability Day is an annual global event (held on the second Thursday in November) focused on usability and human-centred design through talks and local meetups worldwide.

- Teenage Mutant Ninja Turtles is a pop culture franchise whose villain Krang operates a mechanical body, used here to illustrate an AI “harness” (system) versus the model (brain).

- Pac-Man is a classic 1980 arcade game used as a metaphor for tokens as the “bits” an AI system consumes and produces.

Connect with the Guests

- Llewyn Paine, founder and consultant, Llewyn Paine Consulting

- Priya Krishnan, cofounder and COO of Strella

How to AI UXR is supported by Strella, an AI-powered customer research platform that partners with you to build, moderate, and synthesise interviews, allowing you to go from question to actionable insights in just a few hours.

Nam Pham is a senior UX researcher at DoorDash, where they use mixed-methods research to build zero-to-one products. Lately, Nam’s been building dining out as a new category and shaping affordable DoorDash dining. They’re also developing research evaluation practices, making sure that users—in all their messy, human complexity—stay front and centre as the team builds AI and large language model (LLM) products. Before DoorDash, Nam led research on homeowner and seller products at Realtor.com. They studied at Parsons School of Design, where they developed a deep belief in participatory design.

How to AI UXR is supported by Strella, an AI-powered customer research platform that partners with you to build, moderate, and synthesise interviews, allowing you to go from question to actionable insights in just a few hours.

In This Conversation

AI research systems are becoming increasingly easy to build, but how do you ensure those systems are reliable, properly maintained, and deployable beyond the person who built them and their local drive? In this episode, Nam Pham shares a powerful AI research system—one of the most advanced I’ve seen—that’s now part of DoorDash’s core UX infrastructure.

In this episode, Nam demos their research workflow in Cursor and a study plan for a specific delivery question. This system helps DoorDash researchers, designers, and product managers move through the research process, from scoping and planning through instrument design, survey programming, analysis, and sharing.

We talk about how Nam and their DoorDash colleagues use a more traditional tool stack, combined with skills, harnesses, hooks, scripts, connectors, and an internal “skills marketplace” to share and update AI skills that help teams do research.

The primary takeaway from this episode is this: AI might enable you to build solo, but building in collaboration with engineering, design, analytics, and others will enable you to deliver systems that operate beyond one person and their local machine.

The How to AI UXR Map

This series builds on the insights shared in the How to AI UXR map, a five-page map that charts key trends, helps you pinpoint your AI maturity level, and offers practical, real-world applications you can adapt to your research systems.

Download the Map

In this episode, we cover:

- How DoorDash turned individual AI experiments into shared research infrastructure

- What it takes to make an agentic research system usable by researchers, designers, and product managers

- Why a useful AI skill is closer to an operating procedure than a prompt

- How internal knowledge from research repositories, company systems, and Slack can shape research planning

- Why Markdown, reference files, Python scripts, and deterministic code help make LLM outputs more reliable

- How AI can programme a Qualtrics survey while the researcher keeps working, and where human review still matters

- How an internal “skill marketplace” for sharing and updating AI skills supports adoption, maintenance, and shared standards

- How collaboration with engineering, design, analytics, and quantitative research changed what was possible

- Why Nam argues for patience: teach the agent, test it with colleagues, observe where it fails, and continuously and collaboratively improve the system

Partway through the episode, as a leader building one of the tools these researchers are exploring and using, Priya Krishnan, the cofounder and COO of Strella, shares her take on the conversation.

Things Referenced

- Cursor, an AI coding agent

- Claude Code, another AI coding agent

- Glean, an AI platform for work

- MCP (Model Context Protocol) connectors are universal, plug-and-play software bridges that allow AI agents to safely access external tools, databases, and third-party applications

- Markdown is a lightweight plain-text formatting language that uses simple symbols like #, *, and - to structure documents. It’s relevant to AI because it bridges human intent and machine processing, allowing LLMs to easily understand, organise, and output complex information without wasting computing power.

- Python is a computer programming language that’s used to give clear instructions to computers. It acts like a translator, turning English-like words into a format that a machine can easily understand.

- SQL (Structured Query Language) is the standardised programming language used to communicate with, manage, and retrieve data from relational databases.

- AI harnesses are the surrounding software infrastructure, such as tool execution, memory, and safety guardrails, that safely control an AI model and enable it to autonomously execute multi-step tasks.

- AI coding hooks are automated, user-defined scripts that trigger at specific points in an AI assistant’s workflow to enforce guardrails, run tests, or format code.

- Agent journaling is the process by which an AI coding agent maintains a continuous, detailed log of its internal reasoning, tool executions, and step-by-step progress to help developers audit, debug, and understand its decision-making and workflow.

- Skill chaining is the process of an AI agent sequentially linking multiple specific capabilities or tools together, using the output of one action as the input for the next to accomplish a complex, multi-step goal.

Connect with the Guests

- Nam Pham, Senior Researcher at DoorDash

- Priya Krishnan, cofounder and COO of Strella

How to AI UXR is supported by Strella, an AI-powered customer research platform that partners with you to build, moderate, and synthesise interviews, allowing you to go from question to actionable insights in just a few hours.

Daniel Gottlieb is the Head of Research Operations for Microsoft’s CoreAI, where he builds scalable infrastructure for complex UX workflows. Leveraging a PhD background in scientific research logistics and animal behaviour from UC Davis, he focuses on optimising qualitative data management and integrating AI tools into modern research practices. He’s a prominent voice in the field and was featured in the ResearchOps 2.0 documentary series, which explores the past, present, and future of research operations.

How to AI UXR is supported by Strella, an AI-powered customer research platform that partners with you to build, moderate, and synthesise interviews, allowing you to go from question to actionable insights in just a few hours.

In This Conversation

In this conversation, Daniel Gottlieb shares how he moved from feeling behind on AI-enabled coding to building invaluable tools for research operations. What began as a series of personal learning projects—messing around, even—became a set of working research systems: a lab booking app, an interactive budget presentation that will blow your mind, and an AI-powered research library for Microsoft CoreAI.

The How to AI UXR Map

This series builds on the insights shared in the How to AI UXR map, a five-page map that charts key trends, helps you pinpoint your AI maturity level, and offers practical, real-world applications you can adapt to your research systems.

Download the Map

We cover:

- How Daniel went from “I’m not a coder” to building invaluable ResearchOps tools

- Why a homemade video game starring his dog, Tootsie, became a low-stakes way to learn professional AI-building skills

- How he vibe-coded a lab booking app for Microsoft CoreAI, complete with an arcade Easter egg

- Why ResearchOps teams may not always need to wait for engineering bandwidth or a third-party tool to fix everyday workflow problems

- How Daniel turned a budget conversation into an interactive story about researchers, customers, recruiting spend, and business impact

- What happened when he helped his father transform 24,000 astronomy observations into a searchable deep-sky website

- How that personal project became the foundation for an AI-powered research library at work

- Why the library is designed to surface evidence, caveats, and gaps—not false confidence

- What AI means for the future of ResearchOps: more making, more maintenance, and a very different kind of job

Partway through the episode, Priya Krishnan, the cofounder and COO of Strella, shares her take on the conversation.

Connect with the Guests

- Daniel Gottlieb, Head of Research Operations for Microsoft’s CoreAI

- Priya Krishnan, cofounder and COO of Strella

Things Referenced

- Microsoft CoreAI

- Visual Studio Code (VS Code)

- GitHub Copilot CLI

- Steve Gottlieb’s Deep Sky

How to AI UXR is supported by Strella, an AI-powered customer research platform that partners with you to build, moderate, and synthesise interviews, allowing you to go from question to actionable insights in just a few hours.

Michelle Bejian Lotia is a staff experience researcher at Ramp, where she’s been building AI-powered research tools to manage research intake, orchestration, and coaching. Over more than twenty years, Michelle has built and led research teams at Asana, Zapier, and Trainline, establishing voice-of-customer programmes and insights infrastructure. Throughout her career, she’s been driven by one question: How do you bring customers closer to the people building products—and make it scalable? Michelle holds a master’s degree in information from the University of Michigan.

How to AU UXR is brought to you by Strella, a customer research platform that uses AI to run in-depth interviews and generate actionable insights in just a few hours.

In This Conversation

In this conversation, hosted by Kate Towsey, Michelle shares how deadline pressure, combined with the possibilities of AI, led to one of the most compelling examples I’ve seen of AI being used not just to speed up research but also to uplift the craft. Michelle built a transcript-based coaching system that evaluates research calls, provides “tough but fair” feedback, and helps ninety non-researchers improve their craft within minutes of a research session. In this episode, Michelle will walk you through the tool, so make sure to watch closely.

This series builds on the insights shared in the How to AI UXR map, a five-page map that charts key trends, helps you pinpoint your AI maturity level, and offers practical, real-world applications you can adapt to your research systems. → Download the map. 

In this episode, we cover:

* How Michelle built the UXR Interview Coach in a matter of hours

* How the system scores research interviews against a custom-built rubric and delivers private, actionable feedback in Slack

* Why “tough but fair” feedback lands differently when it is timely, specific, and evidence-based

* What the team learned by analysing research quality across roles, call types, and quarters

* How the rubric evolved as the system encountered more nuanced customer conversations

* Why AI’s real value for research may be in enabling new systems, not simply automating old workflows or augmenting analysis

* Michelle’s advice for researchers experimenting with AI: start with one painful or underperforming area and make it better one step at a time

Partway through the episode, Priya Krishnan, the cofounder and COO of Strella, shares her take on the conversation as a leader building one of the tools that these researchers are exploring and using.

Connect with the Guests

* Michelle Bejian Lotia, Staff UX Researcher at Ramp

* Kate Towsey, founder of The ResearchOps Review

* Priya Krishnan, cofounder and COO of Strella

How to AU UXR is supported by Strella, a customer research platform that uses AI to run in-depth interviews and generate actionable insights in just a few hours.