"Volound team on Alpaca AI Trading Agents Hackathon Hackathon"

Team Idea

THESIS is an autonomous options-volatility desk that must convince itself before it trades — and can be cross-examined while it does. A Rust intelligence layer streams option chains across a curated liquid universe, fits arbitrage-consistent implied-volatility surfaces continuously, and hunts one specific class of edge: mispriced event variance around earnings, plus the systematic variance-risk-premium harvest that gives the equity curve its base load. A C++ pricing kernel behind a disciplined FFI boundary does the numerical hot work — vectorized Black-76, full greeks, rational-guess IV inversion, American-exercise fair value. A Gaussian-HMM regime filter gates which sub-strategies may deploy risk; a contextual Thompson-sampling allocator moves capital between them as their theses get graded — the agent measurably learns during the judging window. Every position opens with a machine-readable thesis (quantified expectation, invalidation predicates, exit plan) and closes withTaylor-decom

Team Members

Cade Miller

cademiller23280

AI ML Engineer

Submission

THESIS: options volatility engine

An autonomous options agent that fits arbitrage-free volatility surfaces to live SPY chains, detects where the market disagrees with its own model, and refuses any trade whose edge fails to clear measured execution cost.