AI Engineer · Agentic Systems · MCP · RAG
Building autonomous systems that bridge software engineering and artificial intelligence.
I'm an AI Engineer specialized in agentic workflows and autonomous systems.
My background in Frontend and UX engineering gives me a product-first lens when architecting AI, I care about systems that actually work in production, not just impressive demos.
My current focus: building agents with single responsibility, decoupled by design.
- MCP Servers — custom tools and internal API integrations for AI agents
- Voice AI — outbound calling agents with ElevenLabs, integrated with internal databases
- Context Engineering — moving beyond naive RAG to retrieval systems that actually reason
- Multi-agent Orchestration — single-responsibility agents, composed sequentially, orchestrator owns the state
| Project | Description | Stack |
|---|---|---|
| memoryClaw | Universal Memory & Context Engine as an MCP Server — gives AI agents long-term memory across sessions | TypeScript MCP ⭐ 7 |
| agent-ai-typescript | AI Agents built with LangChain and LangGraph in TypeScript | TypeScript LangChain LangGraph ⭐ 3 |
| sdk-apps-openai | Collection of AI applications with OpenAI Apps SDK and MCP integrations | TypeScript OpenAI MCP |
AI & Agents
Development
Patterns & Protocols
MCP RAG ReAct Chain of Thought Tree of Thought Prompt Engineering
I write about AI engineering, agentic architecture, and what actually works in production.
- How much does it cost to run +4,900 analyses with an OCR Agent?
- In 2026, the RAG we use today will be considered "dumb"
- Every AI Engineer needs to know this — Context Engineering
→ Follow on LinkedIn for weekly content on agentic systems.
"Agents with single responsibility, decoupled by design. State is the orchestrator's job — not theirs."