2026C++, market microstructure, limit order book, ITCH replay, Avellaneda Stoikov, benchmarking, queue position
result3.7M events/sec benchmark, 12,423 QQQ ITCH messages, paired 30 seed statistics, and queue diagnostics.
selected inventory risk and risk adjusted improvement under hand chosen flow; broad PnL dominance did not survive statistical checks.
status: completebenchmark: 3.7M events/secreplay: 12,423 QQQ ITCH messagesconfidence: high
read project2025Python, algorithmic trading
result#194 algo, top 1.4% of 18,800 teams.
the hard part was adapting strategy across rounds under sparse feedback.
status: completeartifact: public leaderboardrank: #194 algo
read project2026honors thesis, Bayesian inference, regime detection, HMM, CUSUM, CPPI
resultBayesian regime detection, HMM plus CUSUM, CPPI drawdown control, and S&P 500 crisis validation.
the hard part was accepting that a well trained Bayesian agent can still stay wrong too long.
status: completeartifact: thesis PDFconfidence: high
read writing entry2025PyTorch, neural SDEs, rough volatility
resultResearch prototype with repo; no public performance number claimed.
the hard part was making path structure matter without overclaiming the metric.
status: research prototypeartifact: GitHub repoconfidence: medium
read project2026Python, Kalshi API, DuckDB, scikit learn, XGBoost
result122 tests, 85% coverage, documented API limits.
the hard part was proving which historical data did not exist.
status: framework completetests: 122 passingcoverage: 85%confidence: high on engineering, documented data limits
read project2026Python, NumPy, SciPy
resultOut of sample Sharpe near 0.7 after fees.
the hard part was reporting the modest number after fixing the inflated one.
status: completeresult: out of samplescope: single asset and one pairconfidence: medium
read project