Blog
Notes on systems, tradeoffs, technical judgment, and shipping work that stays understandable.
I read the Model Context Protocol spec last week and what surprised me wasn't the protocol itself. It was how clearly it removes the "how do I connect this" question from every...
I keep coming back to the moment a delivery pipeline stops being invisible infrastructure and starts being a product feature. It happens when the user's trust in the system is...
I keep coming back to how async Python changes what I can ship, not just how fast it runs. The new @app.vibe() decorator in FastAPI makes this concrete. It is not about...
A small FastAPI dependency bump can expose how much service boundaries are leaking implementation details.
I usually care more about what a bug fix reveals than the fix itself. The FastAPI 0.135.1 release notes mention a fix to "avoid yield from a TaskGroup, only as an async context...
I keep coming back to how human-in-the-loop changes what I can ship, not just how safe it makes the system. The practical question for me is where to draw the line between...
I keep coming back to this: the moment you can predict a change's impact is the moment legacy code stops being legacy. It is not about perfection. It is about predictability....
I read migration plans all the time. The ones that fail share one thing: the analysis and execution drift apart until the document is just a historical artifact nobody trusts....
I usually care more about what a system does than what it's called. The part I would watch in any modernization effort is how teams map dependencies before they talk about...
I built a self-healing deployment pipeline for an AI agent last month. The idea was simple: after every deploy, it detects regressions, triages whether the change caused them,...
I usually care more about what shipping less does to my system's shape than to my product backlog. When I pushed back on adding another auth provider last sprint, it wasn't...
I keep coming back to how we judge AI systems by demo polish instead of operational trust. The systems that look least impressive in a screenshot are often the ones that improve...
A lot of engineering waste comes from trying to look complete before the system is actually clear.
The problem with many AI features is not the model. It is the missing boundary around the model.
Modernization gets easier when the codebase stops being a mystery and starts becoming a map.