Backend Engineer · Spec-driven Builder · Systems & AI Research
I design and build systems from specs, and validate them through experiments and AI-driven automation.
- Spec-driven development (reduce ambiguity → executable systems)
- Hypothesis-driven experimentation (validate before scaling)
- AI agent systems & automation pipelines
- Distributed systems & data processing
- Real-world decision systems (environment → action modeling)
PRD-driven macOS and windows system monitor
- AI-assisted development (PRD → implementation)
- Rust + Tauri desktop architecture
- State diff + notification pipeline
Multi-stage LLM capability framework
- RAG over personal codebase (Spring AI + PgVector)
- Tool-using automation agents (CLI / function calling)
- Multi-agent orchestration (LangGraph-style)
- Recursive self-improving memory system
Weather-based cycling outfit recommendation system
- Environmental condition → action decision modeling (weather → clothing)
- Rule-based + data-driven recommendation engine
- Real-time condition handling (temperature, wind, precipitation)
- Lightweight decision support system for cyclists
AI agent systems · Data pipeline engineering · Automation-first backend design · Decision systems
Turning ambiguous real-world conditions into executable systems.