Hugo Bowne-Anderson, Independent Data & AI Scientist, joins us to tackle why most AI applications fail to make it past the demo stage. We'll explore his concept of Evaluation-Driven Development (EDD) and how treating evaluation as a continuous process—not just a final step—can help teams escape "Proof-of-Concept Purgatory." How can we build AI applications that remain reliable and adaptable over time? What shifts are happening as boundaries between data, ML, and product development collapse? From practical testing approaches to monitoring strategies, this episode offers essential insights for anyone looking to create AI applications that deliver genuine business value beyond the initial excitement.

Hugo Bowne-Anderson, Independent Data &amp; AI Scientist, joins us to tackle why most AI applications fail to make it past the demo stage. We'll explore his concept of Evaluation-Driven Development (EDD) and how treating evaluation as a continuous process—not just a final step—can help teams escape "Proof-of-Concept Purgatory." How can we build AI applications that remain reliable and adaptable over time? What shifts are happening as boundaries between data, ML, and product development collapse? From practical testing approaches to monitoring strategies, this episode offers essential insights for anyone looking to create AI applications that deliver genuine business value beyond the initial excitement.

Catalog and Cocktails is the honest, No-BS, non-sales-y podcast about data and analytics. This is your unfiltered chat about everything interesting in data and analytics. Join Juan Sequeda and Tim Gasper to explore emerging topics and hear from thought leaders, practitioners and executives across the data and analytics space.

Catalog and Cocktails is the honest, No-BS, non-sales-y podcast about data and analytics. This is your unfiltered chat about everything interesting in data and analytics. Join Juan Sequeda and Tim Gasper to explore emerging topics and hear from thought leaders, practitioners and executives across the data and analytics space.

Catalog & Cocktails: The Honest, No-BS Data Podcast

How to Get Out of AI Proof-of-Concept Purgatory with Hugo Bowne-Anderson

Can't believe it's already been 6 years. Thank you to our amazing guests and specially our listeners. In this season finale episode, Juan and Tim rant about the honest no-bs discussions they've had in 2026 and what are the topics they are looking forward to cover in the next season. 

Can't believe it's already been 6 years. Thank you to our amazing guests and specially our listeners. In this season finale episode, Juan and Tim rant about the honest no-bs discussions they've had in 2026 and what are the topics they are looking forward to cover in the next season.&nbsp;

6th YEAR SEASON FINALE: Juan and Tim Rant

Jason Doerr has spent years watching governance programs undermine themselves by cataloging everything without a use case, naming data stewards who have nothing to actually do, and building central teams that become blockers instead of enablers. In this episode, he walks through what pragmatic governance actually looks like: start with use cases, give stewards real work to action on, and let the central team set principles rather than police behavior. He also digs into PADU (Preferred, Acceptable, Discouraged, Unacceptable) as a practical roadmap framework, how LLMs can accelerate semantic layer creation without generating vanity metrics, and why the governance operating model is shifting toward agentic management ... whether the governance community is ready for it or not.

Jason Doerr has spent years watching governance programs undermine themselves by cataloging everything without a use case, naming data stewards who have nothing to actually do, and building central teams that become blockers instead of enablers. In this episode, he walks through what pragmatic governance actually looks like: start with use cases, give stewards real work to action on, and let the central team set principles rather than police behavior. He also digs into PADU (Preferred, Acceptable, Discouraged, Unacceptable) as a practical roadmap framework, how LLMs can accelerate semantic layer creation without generating vanity metrics, and why the governance operating model is shifting toward agentic management ... whether the governance community is ready for it or not.

Pragmatic Use-Case-Driven Data Governance with Jason Doerr

This is the takeaway episode with Jason Doerr who has spent years watching governance programs undermine themselves. He walks through what pragmatic governance actually looks like and digs into PADU (Preferred, Acceptable, Discouraged, Unacceptable) as a practical roadmap framework. 

This is the takeaway episode with Jason Doerr who has spent years watching governance programs undermine themselves. He walks through what pragmatic governance actually looks like and digs into PADU (Preferred, Acceptable, Discouraged, Unacceptable) as a practical roadmap framework.&nbsp;

How to Get Out of AI Proof-of-Concept Purgatory with Hugo Bowne-Anderson

In 1 playlist(s)

Catalog & Cocktails: The Honest, No-BS Data Podcast

Social links

Follow podcast

Recent clips