Hi
I build applied AI and product systems for messy information workflows: RAG, document intelligence, agent workflow state, payout correctness, and backend/product infrastructure.
- Stateframe - file-first state ledger for long-horizon agent workflows.
- PayRail - payout correctness demo with idempotency, ledger accounting, row locks, and state transitions.
- AI Systems Notes - technical writing on RAG, agents, document AI, retrieval, and backend correctness.
| Project | What it proves | Stack / Focus |
|---|---|---|
| Stateframe | Agent workflow state, handoff packets, inspectable task ledgers | TypeScript, CLI, JSON Schema, agents |
| PayRail | Backend correctness for payout flows and double-spend prevention | Django, PostgreSQL, React, Docker |
| AI Systems Notes | System-design thinking around applied AI and retrieval | RAG, document AI, evaluation, technical writing |
| Portfolio | Recruiter-facing project surface | React, TypeScript, Tailwind |
VRAG, bus-fleet, and Cybersec_Exp are public learning/prototype repos. My main proof-of-work is Stateframe, PayRail, AI Systems Notes, and Portfolio.
