I’m a software engineer who has spent a career building and supporting large, real-world systems where correctness matters and failures have consequences. I like the work that sits between theory and production: the place where design meets constraints, and where “it works” has to mean more than “it ran once on my machine.”
In the AI era, the opportunity is obvious: a single engineer can get far more leverage than before. The risk is just as obvious: plausible output can quietly become incorrect output, and “fast” can become “fragile” if we treat language models like magic instead of tools.
My approach is straightforward:
- Use LLMs to compress time, expand exploration, and improve clarity.
- Keep engineering discipline intact: specs, tests, review, observability, and rollback paths.
- Treat reliability as a design requirement, not a postscript.
- Share what I learn so teams improve together, not in isolated pockets.
I write regularly about the practical side of AI-augmented engineering: how to keep quality high, how to reason about tradeoffs, and how to build workflows that survive real constraints like deadlines, load, ambiguity, and human attention.
If you’re here because you typed my domain, welcome. If you’re here because you’re curious about how engineering changes when “thinking” becomes infrastructure, you’re in the right place.