Edward M.

Senior full-stack AI engineer. RAG systems, AI agents and n8n automation, built to survive production. $154,000+ earned, 100% Job Success, Top Rated Plus. Next.js, TypeScript, Python. Most AI projects die between the prototype and production. The demo works, then it hallucinates on real documents, costs 10x at scale, or nobody can maintain it. I build the version that survives that: retrieval that returns the right chunk, agents that fail safely, and infrastructure your team can actually own. Send me your use case and I'll reply with a free Loom on how I'd approach it and a ballpark cost. No call required. WHAT I BUILD 🔍 RAG over your own data. Hybrid search and reranking, chunking that respects document structure, and an evaluation set so you can measure accuracy instead of guessing at it. LangChain, LangGraph, pgvector, Pinecone. 🤖 AI agents that take real actions. Server-side authorisation per tool, retries, cost caps, and human approval on anything irreversible. Not a chat window that returns text. ⚙️ n8n automation replacing manual ops work. Error handling and alerting on every node, idempotent re-runs, environment separation, documented branches. 💻 The full-stack product around it. Next.js, TypeScript, Node.js, Python, Supabase, PostgreSQL, Docker, AWS. Auth, billing, dashboards, deploys. TRACK RECORD • 100% Job Success Score | Top Rated Plus • 4,200+ hours delivered on Upwork across 40+ contracts • $154,000+ earned on the platform since 2020 • AI chatbot handling 5,000+ conversations/month, support cost down 30% • SaaS platform serving 10,000+ users, shipped in 6 weeks • Based in London, with full working-day overlap with the EU and mornings with US East HOW I WORK • Fixed scope, fixed price, fixed date, so you know the number before we start • Your code, your infra, your IP. No lock-in, no proprietary black box • Documentation and a Loom walkthrough at handover, so your team can own it without me • A working demo every week, not a status email ALSO WORK WITH RAG, Workflow Automation, API Integration, PostgreSQL, pgvector, FastAPI, LangGraph, MCP, Make, Zapier, CI/CD Message me your use case and I'll send a Loom back within 24 hours.