A prospecting engine you own.
Scrape prospects, score them 0–100 against your ideal customer profile with AI, write a personalized first line for each, and dispatch the outreach — end to end, self-hosted on your stack. No per-seat SaaS, no rented lead lists.
The open-source core does real extraction — live OpenStreetMap queries, SQLite persistence, an audit log — and replaced roughly $350/month of SaaS with about 600 lines of dependency-free Python. The full outreach pipeline is installed per client; the dashboard below runs on simulated data so you can see the flow without me scraping anyone for a demo.
Four stages, one pipeline.
Find the prospects
Targeted extraction from open data sources or commercial scrapers — your niche, your geography, deduped before anything else touches it.
Qualify with AI
Each prospect scored 0–100 against your ideal customer profile, with the reason stated. Rejects are logged, not lost.
Write the first line
A personalized icebreaker per qualified lead, drawn from what the scrape actually found — not a mail-merge token.
Send and track
Qualified leads flow to your outreach tool and your CRM through n8n. Volumes stay inside deliverability limits on purpose.
Drive it yourself.
Simulated data. The CSV export is real — try it.
Start small, keep everything.
A pilot install is a fixed $1,500 — one pipeline, shipped in two weeks, documented and handed over. If it earns its keep, the retainer takes it from there.