Data Science · AI Engineering · Systems Architecture
02
01 — About
I think in systems, not scripts.
Given a hard problem the first thing that forms isn't a function — it's an architecture: the agents, the memory, the data flow, and where it breaks. I design the failure modes before I write the happy path.
multi-agent
RAG
applied ML
02 — Who
Systems that admit what they don't know.
Every model here can be asked why. Retrieval that cites its page, a classifier that shows its reasons, an engineering rig gated against its own claims. I'm building toward AI you can put in front of an expert without embarrassment — because a confident wrong answer costs more than no answer at all.