Machine Learning Engineer @ LPL Financial · Prev. Quantitative Developer @ QuantConnect · ex-President @ Triton Quantitative Trading, UC San Diego
B.S. Computer Science, UC San Diego (2026) · LinkedIn · QuantConnect
I build low-latency trading and research infrastructure, quantitative signal pipelines, and the tooling that sits between them.
Stack: Python · C++ · C# · PyTorch · NumPy/Pandas · SQL · Apache Arrow/Kafka/Spark/Airflow · Redis · AWS · Docker · Kubernetes
Spectral Tick-Flow Signal: proprietary Fourier-transform pipeline detecting systematic execution algorithms across 500 US equities since 2009; 2.5x faster parsing via Apache Arrow, 5.8x faster via multiprocessing. Research →
FinMamba3: native C++ (pybind11) backtesting engine for prediction-market limit-order-book data; releases the GIL for concurrent runs, parity-verified byte-identical against the original Python engine.
DataHacks 2026 Backtester: event-driven backtester for binary prediction markets (5k+ LOC, 93 tests); order-book-walking fills with liquidity depletion, T+1 latency, stale-quote rejection across 8,466 markets.
Architecting the Agent Orchestration core of QuantConnect's AI framework · Mentoring TQT student researchers (LPPLS bubble detection, NLP filing-language signals, CNN-LSTM volatility forecasting) · Lecturing CSE 198: Introduction to Quantitative Finance at UCSD
AI Orchestration and Agents at scale.
📫 ruosuna@ucsd.edu · Trilingual: English, Spanish, Portuguese | Learning: Mandarin (Simplified)



