PinSight
Infers the risk-neutral density of SPY same-day options and runs a scheduled live paper-trading loop. Every decision is a pure function of an as-of timestamp, and a test proves past decisions cannot change when future data is appended.
A research and paper-trading platform for Polymarket and Kalshi binary contracts, where the price is never mistaken for the probability.
Key facts
64 tests
with continuous integration
5 min / 30 s
market and top-orderbook polling cadence
Zero size
for any bet with no evidence of edge
How it works
- Kelly sizing on an empirical-Bayes win probability. The prior is the gambler's-ruin probability of hitting the target before the stop, under which Kelly's expected value is exactly zero; realized trades update it.
- Venue-outcome settlement clears positions stuck on resolved markets whose orderbooks no longer exist.
- Deterministic no-lookahead replay over recorded orderbooks audits every change. One audit cleared a suspected code regression and traced the losses to a daemon failure and an edgeless strategy, which was retired.
Built with
Python, Polymarket and Kalshi APIs, pandas, Parquet, pytest, GitHub Actions
What it does not show
- Paper trading only. No trading performance is claimed; the engineering and the audit discipline are the point.
Tanishk Yadav
MS Financial Engineering at New York University, graduating May 2027, and looking for full-time quantitative research roles. My master's thesis measures, in real time, when US debt dynamics turn self-reinforcing.
Related
Related work
RateWalk: Forecasting Fed Decisions
Naive inflation-regime conditioning overfits; empirical-Bayes shrinkage fixes it in the US, UK and Germany, and a blend with a market proxy reaches log-loss 1.056 and a 72.7% hit rate.