High Signal: Data Science | Career | AI
Delphina
Welcome to High Signal, the podcast for data science, AI, and machine learning professionals.
High Signal brings you the best from the best in data science, machine learning, and AI. Hosted by Hugo Bowne-Anderson and produced by Delphina, each episode features deep conversations with leading experts, such as Michael Jordan (UC Berkeley), Andrew Gelman (Columbia) and Chiara Farranato (HBS).
Join us for practical insights from the best to help you advance your career and make an impact in these rapidly evolving fields.
More on our website: https://high-signal.delphina.ai/
High Signal brings you the best from the best in data science, machine learning, and AI. Hosted by Hugo Bowne-Anderson and produced by Delphina, each episode features deep conversations with leading experts, such as Michael Jordan (UC Berkeley), Andrew Gelman (Columbia) and Chiara Farranato (HBS).
Join us for practical insights from the best to help you advance your career and make an impact in these rapidly evolving fields.
More on our website: https://high-signal.delphina.ai/
Episodes
Mentioned books
22 snips
Mar 19, 2026 • 1h 4min
Episode 36: AI and the Judgment Problem in Data Science
Andrés Bucchi, a senior data leader at LATAM Airlines focused on scaling data teams and secure AI deployment, and Dawn Woodard, a distinguished engineer with experience building experimentation and analytics platforms, discuss AI’s impact on analytics. They cover the source-of-truth challenge, verifiable outputs and catalogs, agent-first tooling, experimentation bottlenecks, and how roles shift toward validation and trustworthy AI workflows.
16 snips
Mar 5, 2026 • 55min
Episode 35: Beyond Online Experimentation: Generative Software That Optimizes Itself
Martin Tingley, experimentation leader at Microsoft and former Netflix experimentation head, explains why humans are the bottleneck in testing. He outlines a five-level maturity framework moving from basic tests to AI that generates and refines product variants. Topics include parameter optimization, automated explore-exploit systems, generative AI closing the loop, and how experimentation informs strategy and org roles.
Feb 10, 2026 • 46min
Episode 34: Duolingo and the Future of Personalized Education with AI
Bozena Pajak, VP of Learning at Duolingo and former learning scientist in linguistics and cognitive science, discusses AI’s evolution at Duolingo. She covers personalized difficulty models, generative AI characters for low-stakes speaking practice, agentic workflows to scale content, and a shift toward thematic lenses for deeper personalization.
49 snips
Jan 27, 2026 • 1h
Episode 33: Why Your AI Product Will Be Obsolete in Six Months (And What To Do About It)
Ben Stancil, writer and Mode co-founder known for sharp essays on AI and product design. He argues AI shifts work from doing to obsessive polish. He warns building heavy AI harnesses now risks rapid obsolescence. He envisions communication mediated by shared repositories and predicts messy prototypes may become useful specs for clean rewrites.
22 snips
Jan 13, 2026 • 42min
Episode 32: The Post-Coding Era: What Happens When AI Writes the System?
Nicholas Moy, former Head of Research at Windsurf and now at Google DeepMind, shares insights on the transition from traditional coding to a new agentic development era. He discusses how Windsurf pivoted from static autocomplete to successful agentic coding products. Moy emphasizes the importance of real user data for training models and warns startups to 'disrupt themselves' to stay relevant. He believes that true defensibility now lies in brand reputation and community engagement, as agents take on more autonomous roles in software development.
9 snips
Dec 30, 2025 • 47min
Episode 31: Why Data Governance In Your Org is Broken (And How to Fix It)
Cara Dailey, Senior Data Leader at Early Warning and former CDO in financial services, shares her insights on data governance. She champions a pragmatic 'progress over perfection' mindset and argues for treating data as a strategic product rather than a monolithic system. Cara emphasizes the need for ownership, quality, and clear stewardship roles in data management. She also highlights the importance of security and cautious AI adoption, envisioning a future where conversational AI enhances data governance and usability.
24 snips
Dec 11, 2025 • 50min
Episode 30: The AI Paradox: Why Your Data Team’s Workload is About to Explode
Chris Child, VP of Product at Snowflake and author of an MIT Technology Review report, dives into the paradox of AI increasing data teams' workloads. He discusses how data engineering is shifting from backend functions to vital business strategy. Chris highlights the need for data engineers to think like product managers and the challenges faced with LLMs lacking business context. He emphasizes the importance of investing in foundational governance and exploratory experimentation to navigate rapid AI changes effectively.
13 snips
Nov 28, 2025 • 49min
Episode 29: Why AI Adoption Fails: A Behavioral Framework for AI Implementation
Liz Costa, Chief of Innovation and Partnerships at the Behavioural Insights Team, shares insights on AI adoption through the lens of behavioral science. She discusses the crucial role of understanding human behavior in fully utilizing AI's potential. Costa introduces a triad of motivation, capability, and trust as barriers to adoption, while emphasizing the importance of reframing AI deployment to alleviate skepticism. She also highlights the need for organizational experimentation to uncover valuable use cases, ensuring deep integration rather than mere automation.
120 snips
Nov 13, 2025 • 51min
Episode 28: From Context Engineering to AI Agent Harnesses: The New Software Discipline
Lance Martin, a machine learning engineer at LangChain, dives into the evolving landscape of AI engineering. He emphasizes the importance of context engineering and how traditional ML rules are becoming obsolete. The conversation covers why adaptable systems thrive, the architectural advantages of 'agent harnesses,' and the shift towards in-app user feedback for evaluating AI systems. Lance also shares insights on balancing autonomy in agents with human oversight and techniques for managing costs and performance in complex AI tasks.
28 snips
Oct 30, 2025 • 42min
Episode 27: Why Your Data Team Doesn't Have a Seat at the Table (And How to Earn It)
Paras Doshi, Head of Data at Opendoor and former Amazon data leader, shares insights on building effective data functions. He discusses transforming fragmented analytics into a centralized asset, emphasizing the importance of having a 'seat at the table' for data teams. Paras argues that AI is set to create the '100x individual contributor,' enabling significant productivity boosts. He also covers the balance between batch and real-time ML, the role of data in strategic business decisions, and the essentials for a successful data career.
