LeadOps Autopilot: Sourcing and Scoring US Clinic Leads from Public Data

Description and use

LeadOps Autopilot is an automated pipeline that sources US clinics from public registries and maps, enriches with website data, scores each lead with transparent reasons, and outputs a human‑reviewed CSV before any outreach is sent.

Problem

Clinic lead lists are often stale, incomplete, or irrelevant because they only confirm existence without indicating operational fit for lead‑intake automation.

Outcome

The system provides provenance‑rich, score‑based clinic leads with transparent reasoning, enabling cheap iteration, compliance by construction, and visible failure modes, while human approval gates outbound outreach.

Workflow

• Collect clinic data from CMS/NPPES and OpenStreetMap Overpass API
• Enrich with website discovery via Google Places API
• Fetch each clinic’s homepage and score based on fit and negative signals
• Generate pain hypothesis and draft outreach copy
• Segment leads by vertical and batch by fit tier
• Human reviews CSV and approves high‑fit rows before any outreach is sent

Evidence

Most lead lists are bought, stale, and sprayed. Ours is none of the three — and the interesting part is not the scraping.

Source

The scorer (<code>leadops.py</code>) fetches each clinic's homepage and scores what it finds there.

Source

The MVP stops before sending.

Source