Notes on building reliable software with AI in the loop.
Flagship Case Study
A lightweight governance layer for AI-assisted development that makes system intent explicit, auditable, and durable across agent sessions.
Tools: MCP (Model Context Protocol), Claude Code, JSONL Event Log, Markdown Intent Registry, Agent Skills
Designing a multi-scenario strength evaluation system inspired by leadership principles and built to distinguish human conviction from synthetic fluency.
Tools: Next.js, TypeScript, Supabase, PostgreSQL, LLM APIs (OpenAI / Gemini), Structured Prompting, Custom Evaluation Engine
Designing a privacy-first, want-driven coordination system using AI-assisted full-stack development.
Tools: Next.js, Supabase, PostgreSQL, Row Level Security, pgvector, OpenAI APIs, ChatGPT, Cursor, GitHub Copilot
Validating AI-assisted full-stack mobile development through a solo-built, production-ready MVP.
Tools: React Native, Expo, Supabase, Postgres, ChatGPT, Cursor, GitHub Copilot
How AI compressed implementation time and exposed design clarity as the true bottleneck.
Tools: React, Next.js, Expo, Supabase, Postgres, ChatGPT, Cursor, Claude Code, Codex