About
AI systems, agentic workflows, and business interfaces.
I build working AI systems across document intelligence, customer support agents, business workflows, and automation platforms — with clear architecture, traceable outputs, and practical business use cases.
Systems I build
Document intelligence & RAG
Source-grounded answers from documents, PDFs, and records.
Customer support agents
Policy-grounded decisions with reasoning logs and escalation.
Automation workflows
n8n pipelines with human approval on consequential actions.
Business workflow platforms
POS, dashboards, and operational systems on structured data.
Transcript & review intelligence
Conversations turned into structured, actionable insight.
Built with production discipline — evidence gates, human approval on consequential actions, and fallbacks designed before launch.
How a system flows
Inputs become reasoning, reasoning becomes traceable output — the same shape across every system I build.
Core technical stack
- Python
- TypeScript
- Next.js
- FastAPI
- LangChain
- LangGraph
- n8n
- RAG
- OCR
- Vector Search
- Structured Outputs
- AI Agents
- Prisma
- Dashboards