I work primarily with machine learning applied to quantitative finance — feature selection via block permutation importance, multi-objective gradient boosting, out-of-sample validation with statistical rigor (bootstrap CIs, effect sizes, significance testing), and end-to-end backtesting infrastructure.

Outside of that, I write performance-critical C — data compressors using context-mixing techniques and cache-friendly parallel sorters for Linux.

I build products end-to-end — SaaS applications, API services, chatbot integrations, and business automation tools. From backend architecture to deployment.

I also build AI-powered developer tooling — CLI tools that integrate LLMs into engineering workflows.

Other interests include game server emulation (C#) and retro game development.

Python · C · TypeScript · C# · LightGBM · Node.js · FastAPI

Open to conversations about quantitative research, ML infrastructure, or systems-level optimization.