AI lead scoring and enrichment engine for 10+ B2B companies

One platform that finds, scores and enriches B2B leads for 10+ agencies and B2B companies.

The clients needed more of the right leads and were still prospecting by hand. I designed and built about 80% of a shared engine that watches social posts, competitor audiences, intent signals and tech stacks, scores every profile against the client's ideal customer, and only then pays for enrichment. Small models make the fit calls and rules handle the rest. Each client is a config file: one wanted quality only (435 qualified leads for €3 on the first run), another wanted volume (7,000+ leads, 28% reply rate).

Waterfall enrichment across 9 providers, re-scoring a lead when new data lands

€0.001 to €0.12 per profile, with spend caps per provider

40% of one client's pipeline from AI-surfaced deals, 25% reply rate

Python and FastAPI on Vercel, 2,644 automated tests