I'm a software engineer who builds systems that turn uncertainty into evidence.

For the past year I've been designing and running a quantitative research platform: an end-to-end pipeline that takes market hypotheses from historical backtesting through forward validation, paper/demo execution, and portfolio allocation — with automatic retirement of ideas that fail. Most ideas fail. The platform exists to prove which ones, and why.

• Market research platform — TypeScript/Node monorepo: backtesting engine, forward-validation paper tapes, live execution against exchange APIs, risk overlays, regime detection, cross-strategy correlation analysis, and a signal-analytics layer that logs every accepted and rejected signal so filters can be audited for what they cost.
• Runway — an AI-assisted job-discovery and application pipeline with automated follow-up sequencing and inbox synchronization.
• Prediction-market research lab — the same validation methodology applied to a different asset class; shared infrastructure, different data adapters.
• AI English tutor — production Vue/AI app for conversational language practice.

Every idea goes through the same lifecycle: hypothesis → historical validation → forward validation → execution verification → allocate or retire. The write-ups I care most about are the negative results — survivorship traps, look-ahead bugs, fill-price artifacts — because that's where the engineering rigor actually lives.

10+ years of full-stack engineering: Vue/Nuxt, Node/TypeScript, Python, e-commerce platforms (Shopify), real-time systems, and AI-integrated products.

📫 Open to remote Senior Software Engineer / Founding Engineer roles — especially research infrastructure, data/AI platforms, and fintech.