I build performance software with rigorous evaluation and realistic assumptions that holds up under real world constraints.
Tools, systems, and ideas that solve problems and actually get used.
I move fast, but I care a lot about getting things right the first time.
Understanding the problem deeply before jumping to solutions.
Hard problems in competitive spaces where performance matters.
A diagnostic framework for stress testing quantitative strategies and identifying fragility under noise, regime shifts, and adverse conditions.
A research-grade quantitative IDE with an agentic MCP server for coding, testing, and iterating multi-asset trading strategies under realistic constraints.
A C++ simulation engine for modeling gravitational interactions with an emphasis on performance, numerical stability, and deterministic behavior.
A classifier that estimates whether an image is AI generated, built with CLIP embeddings and a calibrated supervised model.
Reinforcement learning agents trained in simulation for control, navigation, and decision-making under noisy and competitive environments.
An LLM powered mental health support application focused on reliable conversation flow, state management, and real world user feedback.
