You've seen the demos. An agent builds an app from a single prompt, and the crowd applauds. Then you go back to work and ask your coding agent to add a feature to a solution that's been in production for twelve years. The results look nothing like what you saw in the demo.

AI-native development is no longer a novelty. Most teams now have coding agents in their editors, in their pull requests, and increasingly in their pipelines. But experiences vary widely. Some developers still can't get the results they saw in a keynote to hold up on their own code. Others say they've automated much of their SDLC with AI software factories that take work from backlog to production. Most professional developers shipping real software to real users are somewhere in between, trying to sort out what's real, what's hype, and what's worth their time.

Eric Boyd sits down with three developers who have spent decades building, shipping, and supporting software. All three are putting AI to work in their own code and with the teams they advise:

• Rocky Lhotka is a VP at Xebia, the creator of CSLA .NET and RockBot, and a leading voice on AI-assisted coding. Rocky brings the engineer and architect's view: what AI means for application design, frameworks, and the business logic that has to stay correct.
• Benjamin Day is a fractional CTO, Microsoft MVP, and Azure DevOps and GitHub expert who maintains more than 20 open-source .NET libraries and tools. Ben brings the delivery view: how AI is changing testing, team process, and the path to production, and why the bottleneck that looks technical is so often organizational.
• Bob Ward is a Principal Architect on the Microsoft Azure Data team, a 32-year Microsoft veteran who has worked on every version of SQL Server, and author of the Revealed book series. Bob brings the data view: from writing and tuning queries to building AI-powered applications on SQL Server, Azure SQL, and Microsoft Fabric.

We'll start with a quick level set on where AI-native development stands today. From there, it's an open conversation among seasoned developers about what's working, what isn't, and the advice they'd give their teams, their customers, and themselves. Expect honest answers, some friendly disagreement, and plenty of time for your questions.

What we'll discuss:

• Where AI is delivering real productivity gains today, and where the gains are mostly hype
• Brownfield reality: using agents on large, aging codebases, not just greenfield demos
• Trust and verification: how these developers review, test, and validate code they didn't write themselves
• What's changing in team process, roles, and the shape of a developer's day
• AI and your data: queries, schemas, and AI features built on the databases you already run
• The skills and fundamentals that matter more now, and the ones that matter less
• Mistakes they've made, so you don't have to