The Joe Reis Show

On this episode of The Joe Reis Show, I’m joined by Dan Bennett, Head of Technology for the Enterprise Data Organization at S&P Global.We get into what it actually looks like to build with modern AI coding tools like Claude Code. Not just as a toy, but for writing production-grade C++. Dan walks through how he built an open-source RDF extension for DuckDB, why rock-solid test coverage is non-negotiable when working with LLMs, and why engineering leaders need to keep their hands dirty to understand where this tech is headed. We also react to the breaking news of AWS acquiring DuckDB Labs, talk through the shift from "human-in-the-loop" to autonomous agents running in headless VMs, and dive into why data semantics across organizational boundaries remains one of the hardest - and most important - unsolved problems in our industry.Website: https://nonodename.com/

I discuss the The Mixed Model Arts Manifesto. That's it. That's the show.

Will AI agents solve the data industry's hardest challenges, or are we just amplifying existing chaos?In this episode, I chat with Oliver Laslett, Co-founder and CTO at Lightdash, to explore the frontier of agentic analytics, data modeling, and modern software engineering practices. We dive deep into how AI is shifting daily workflows from writing SQL and DBT models to higher-order system thinking, why the hardest data problems remain human and organizational, and the cultural differences in tech optimism between London and San Francisco.Oliver also breaks down why permissions, curation, and data lineage are still the true bottlenecks, and why high-ownership communication matters more than ever in an era of AI slop.

It feels like everything is converging right now. Data forms and systems are converging. Agile and Waterfall are starting to look strangely compatible. Software engineering, data engineering, and AI engineering are bleeding into each other.
In this episode, I explore why AI is collapsing boundaries we spent decades creating, and why semantics, architecture, specification, and judgment may matter more as a result.
-----------------------
Sponsor: Fivetran
With the rise of AI and agents, having centralized, trustworthy data is the absolute foundation for building tools and training models. Fivetran automates your data pipelines, removing the need to build fragile connectors so your data arrives clean and reliable. By handling the messy infrastructure behind the scenes, Fivetran allows your team to focus on building the future. Visit ⁠⁠⁠⁠fivetran.com⁠⁠⁠⁠ to learn more.

The Joe Reis Show is invitation-only. Unsolicited guest pitches and PR outreach are not considered.

Are massive frontier AI models actually worth the cost?In this episode, I sit down with Daniel Svonava, Founder of Superlinked, to break down where AI is actually heading in production. We dive into why the hype around massive frontier models is shifting toward targeted, small open-source models, and how to orchestrate smart agents without blowing through your budget.Daniel shares how Superlinked is building open-source inference infrastructure, why small models are rapidly closing the performance gap, and how you can engineer faster, cheaper, and more control-driven AI systems.