Mike Betz.

I started my career when Oracle 8i was considered cutting-edge and enterprise software came on CD-ROMs. Over the next two decades I worked across databases, integrations, commerce platforms, and HR systems — building things that had to actually work at scale, under pressure, with real business consequences if they didn't.

I've designed data architectures on Oracle and SQL Server, built integrations on PeopleSoft and SAP Hybris, written more stored procedures than I care to count, and led teams through the kinds of projects where "it'll be fine" is never an acceptable answer.

That background gave me something I didn't fully appreciate until recently: I know what it looks like when technology actually works in production. Not in a demo. Not in a sandbox. In the real world, with messy data, legacy constraints, and executives who need answers on Monday morning.

The Shift

A few years ago I started paying serious attention to AI. Not the hype cycle version — the practical question of what these tools actually do well and where they break down in enterprise environments.

The more I dug in, the more I realized this wasn't another ERP boom or NoSQL wave. The fundamentals are genuinely different. The mental model required to work effectively with LLMs is different from anything I'd learned before, and my instinct to over-engineer and demand deterministic behavior had to be consciously unlearned.

I decided to go deep. Build things. Document what I learn. Share it publicly so others making the same transition have something real to read — not marketing copy or academic theory, but the honest account of an experienced engineer figuring out where the rules changed.

What I'm Doing Now

I work as a Technical Architect at Capgemini, focused on commerce and enterprise platform integration. Outside of that, I'm building AI applications, exploring agentic systems, and writing about what I find at michaelbbetz.com.

This site is entirely built with AI tooling. That's intentional. The best way to understand what these tools can do is to use them to build something real, not just run experiments in isolation.