All projects · AI Agents & RAG
Lean Lead Finder v2 (Gemini Chain + Hunter + Airtable)
An n8n lead-generation pipeline that uses a Gemini LLM chain to invent a list of real companies matching a configurable Ideal Customer Profile,…
An n8n lead-generation pipeline that uses a Gemini LLM chain to invent a list of real companies matching a configurable Ideal Customer Profile, enriches each with contact emails via Hunter.io domain search, then writes qualified (and no-email) leads into Airtable. Runs on demand or daily at 8am.
- AI chain with a Structured Output Parser that forces Gemini to return guaranteed {companies:[{name,domain,description}]} JSON — no downstream defensive JSON.parse
- Config-only 'Define ICP' layer (industry, keywords, location, company size, target roles, max companies) as the single edit point for who to target
- splitInBatches 'One Company at a Time' loop isolates each Hunter call so one 404 doesn't kill the run; Loop Back node re-feeds the batch
- IF-node regex email validation branches into two Airtable writes: qualified leads vs a 'no email' row kept for later retry (dedup/retention logic)
- Domain cleaning (strips protocol, www, path, lowercases) and one-Airtable-row-per-email fan-out; dual triggers (manual + daily schedule at hour 8)
Built with: n8n, Google Gemini / PaLM (LangChain chainLlm), Structured Output Parser (LangChain), Hunter.io Domain Search, Airtable, Manual Trigger + Schedule Trigger, Split Out + Split In Batches loop, IF node (regex), Code nodes