I build AI applications and agent systems.
Most of the time, I am building AI products and agent systems that can actually be used, not just demoed.
I care more about whether something is useful than whether it looks futuristic in a screenshot.
I do not stay inside one stack for long. Python, Go, Java, and JavaScript all show up in my day-to-day work, depending on what helps the product move forward.
I used to spend a lot more time on Kafka and Flink. They are not the center of my work now, but I still know that world well enough to pick it up when needed.
- AI products that solve real problems
- agent workflows, tools, and the backend code around them
- product engineering that turns prototypes into usable software
- side projects that are small, practical, and a little fun
Kafka and Flink are still in the toolbox when streaming or real-time data shows up.