A working space, not a résumé.
I work across financial systems, applied AI, research workflows, and model-driven decision making. None of those is quite a job title, which is the point — the interesting problems tend to sit in the seams between them.
Day to day I work in model risk, covering financial and statistical models across risk, stress testing, and business planning. Earlier I worked in quantitative advisory and model-development-adjacent work. That background is where a lot of my instincts about reliable systems around imperfect models come from — you spend enough time with model inventories and validation and you start seeing the same structural problems everywhere.
Alongside that I build applied AI systems: agents, retrieval, evidence trails, the unglamorous plumbing that decides whether a prototype survives contact with a second run. The two halves inform each other more than people expect. Financial modeling is a discipline of being careful about what you claim and why; that turns out to be exactly the muscle applied AI is missing.
Why this site exists
Most of what I make and think doesn't fit neatly into a paper, a repo README, or a LinkedIn post. This is where the in-between things live — notes developed enough to share but not finished, projects with their status showing, ideas I'm still turning over. It grows slowly and in public, which is the only way I've found to keep a body of work honest.
It is deliberately not a portfolio. There's no pitch here. If something is useful to you, good; that's a nice side effect of writing things down.