The retail investor in 2026 has more information than any pension manager had in 1990 — and worse outcomes. The problem isn't access. It's discipline. Every brokerage app is engineered to make you trade more. Every news feed is engineered to scare you out of a position you should have held for a decade.
We thought: if reasoning models can write a research memo as well as a junior analyst, surely they can also be the steady hand the average investor needs. Not to pick the next moonshot — but to do the unglamorous work of holding, rebalancing, harvesting losses, and ignoring the noise. That's the entire product.
Builds the agent framework that powers Claremont. Computer Science at Harvey Mudd College, with a focus on applied AI and large reasoning models. Multiple early-stage startups under his belt before this — the kind of operator who has shipped a v1 at 3am and learned every lesson the hard way.
Pomona College, Class of 2026. After a stint at a Korean hedge fund, went on to run his own discretionary book — and personally posted a +265% cumulative return over four years. Brings the actual investment process — risk budgeting, position sizing, conviction frameworks — that the agent runs on. The patient half of the partnership.