n8n + Claude AI Agent — production-grade build, 1-2 week delivery

About this Gig

I'll build a production-grade AI agent for you on n8n + Claude API. Not a prototype, not a tutorial — a workflow that runs in production from day one. Typical scope: - CRM-aware AI setter (inbound qualification, follow-up, routing) - RAG over your knowledge base - Webhook orchestration with durable retries - Postgres/Supabase-backed state for resumability Stack: n8n (self-hosted or cloud), Claude API (Opus/Sonnet/Haiku per cost+quality fit), Supabase or Postgres for state, your CRM (HubSpot / GoHighLevel / similar). Optional self-hosted Ollama for cost reduction. Deliverables: - Working workflow in your n8n instance - Scope doc with runbook - Monitoring + alerting wired up - Hand-off video walkthrough 5+ years across automation and AI integration. I run this exact pattern across 3 of my own businesses — 4 n8n instances, 288 synthetic probes/day, zero submissions lost in 4 months of production. I bring the same approach to your project.

Requirements

From you to start: - Brief: what trigger fires the agent, what outcome it should produce, the success metric - Access: sandbox n8n instance (OR I'll spin one up for the build), Claude API key (OR I'll proxy mine for the test mission) - Sample data: 5-10 example inputs the agent should handle (anonymised if needed) - Stakeholder availability: a 30-min kickoff call plus async daily check-ins If scope is unclear, that's fine — I'll do a 30-min discovery call first to scope properly before quoting.

Related Tags