I build products, own systems end to end, and use AI as a native engineering layer to move faster without compromising quality.
- Designing agentic workflows with tool use, MCP, hooks, and reliable handoffs
- Using AI-native engineering in day-to-day product and systems work
- Building production-ready systems with clear architecture, observability, and release discipline
- Shipping products from discovery through design, build, iterate, and scale
- Treating context management and tokenomics as first-class constraints — budget what the model sees, measure cost per outcome, cut waste without losing quality
Native AI loop: Spec → Agent (MCP / tools) → Review → Eval → Ship — with context budgets, prompt hygiene, and token attribution at every handoff.
AI Tools (Daily Drivers)
-191919?style=for-the-badge&logo=anthropic&logoColor=white){kind=icon}
-000000?style=for-the-badge&logo=modelcontextprotocol&logoColor=white){kind=icon}
| Project | What it is |
|---|---|
 |
Plug-and-play AI observability for Copilot, Claude Code, Cursor, and Gemini — attribution, metrics, and quality |
 |
Human-in-the-loop MCP tools + prompt hygiene to cut Copilot premium requests and token waste |
 |
Public community study guide for Claude Certified Architect Foundations |
 |
Local discovery and community product in progress |
 |
Companion community platform in progress |
- Collaborating on AI-native product and agentic system builds
- Sharing architecture patterns for Claude, Gemini, Copilot, Cursor, and MCP
- Mentoring engineers adopting AI-first workflows
