Selected work
Work
01
Market-data operations toolkit
Classifies real CME and Nasdaq notices, resolves everything they touch through a typed dependency graph, and routes each one for review with an explicit confidence and escalation rule — then scores itself on a held-out split frozen before any rule was written, and shows its failures on the page.
Python · D3 v7 · 26 real notices · synthetic estate
Held-out · real notices (n=7)
type accuracy1.00
feed recall0.86
priority accuracy0.43
false clears1
02
IMC Prosperity 4 — global trading competition
Built and tested strategies across market making, order-book microstructure, time-series analysis, backtesting, and game-theoretical decision-making, over a multi-round global event.
Python · backtesting · microstructure · game theory
Result
Top 1.3%
#242 / 18,803 teams · #10 in NL
03
Fullhouse poker AI — international championship
An adaptive bot for 6-max no-limit hold’em, evaluated across 10M+ simulated hands, 100+ agent variants, and 30+ custom opponent models.
Python · simulation · opponent modelling
Result
5th
international · 6-max NLHE
04
Astronomy data project — national informatics competition
Analysed Milky Way stellar structure with Gaia DR3 and ADQL under PhD mentorship, as a high-school computer-science research project.
Python · Gaia DR3 · ADQL
Result
2nd
nationwide · Netherlands
In progress
- Wavelet-transform pairs trading
- ETF tail-hedge risk