Systematic infrastructure for execution, risk oversight, and macro research. Mostly Python, with a few Node.js, Docker, and GraphRAG pieces.
Traders pass a paid evaluation on a sim account funded with live Kalshi and Polymarket prices. Pass the eval, get a funded account, take 80% of the profits. I built the platform end to end: onboarding, eval engine, live market data pipeline, payouts, and the ops dashboard that keeps it all honest.
A multi agent system built on GraphRAG that synthesizes 10,000+ data points into directional signals. A digital twin sandbox runs 500+ stress tests to model portfolio sensitivity, while a Zep Cloud pipeline lifts reporting efficiency by 40%.
A monitoring engine covering ~25 daily trades with real time position reconciliation. Live P&L over WebSocket APIs cut trade discrepancies by 15%, and automated risk controls keep portfolio drawdown under 5%.
Aggregates 25+ global sources for macroeconomic risk indicators. LLM summarization shortens analysis of dense financial reports by 70%, and real time alerts flag high impact central bank policy shifts.
A hierarchical AI framework managing the end to end lifecycle of 30+ systematic financial models. Automated reconciliation between research signals and execution data secures 100% data integrity.
A Python framework automating 350+ daily settlements in high velocity markets. Synchronizes CLOB and Gamma APIs over WebSockets, orchestrated by AI agents that own the full settlement lifecycle without manual intervention.