I build the data systems that make AI agents actually work.
As a Senior Analytics Engineer I design real-time data models, migrate BI platforms, and build the infrastructure that lets LLMs answer questions humans actually care about.
I sit at the intersection of data engineering and analytics, designing the models, tools, and semantic layers that analysts and AI agents use to do their jobs. Right now that means building real-time data infrastructure for an LLM-powered hiring agent, wrangling ClickHouse into doing things I wasn't sure it could do, and thinking hard about where the line is between giving an AI enough structure to be useful and so much structure that you've killed the point of having one.
Before focusing on data, I spent years in chemical process design and safety consulting for chemical and oil refineries analyzing complex systems, identifying failure points, developing hazardous release models and building frameworks to prevent catastrophic outcomes. It turns out data engineering is not much different, just with fewer explosions.
I write about what I learn.
-
If you prescribe every outcome, you've built in every bias you already have. Precision and prescription aren't the same thing. Most teams are conflating them.
-
When an agent starts behaving unreliably, the instinct is to add more instructions. It makes things worse. The fix is precision, not volume.
-
One camp engineers the agent so thoroughly no human has to intervene. The other believes the human-AI collaboration is the product. I'm in camp two.
-
I don't assume it knows everything, and I don't try to make it do everything. I use it as a thinking partner. Here's what that actually looks like.
-
I didn't start in data. The skill set from process safety management consulting transferred more directly than you'd expect.
