Back to Portfolio
2026
GeminiScrapingWhatsAppMaps

AI-Graded B2B Prospect Engine

"Scraped 84K local travel businesses from Google Maps, had an AI model grade every one by how likely it sells bus tickets, and turned it into a 4.6-lakh-prospect field map for the sales team — the engine behind 10K+ verified agents onboarded from zero."

Executive Summary

To launch a B2B agent channel from scratch, we needed to know who to call. I built a collector that scraped travel businesses PIN-code by PIN-code, used Gemini to read each listing and review set and grade it from proven bus seller to cold lead, merged it with public agent directories, and shipped a field-map dashboard where account managers see every prospect by state, city and PIN with live onboarding status. WhatsApp outreach on top of it brought in 10K+ verified agents.

Key Outcomes

  • 84K businesses scraped and AI-graded; 4.6L prospects mapped in total.
  • Field-map dashboard with Google sign-in used by the sales team.
  • 10K+ verified travel agents onboarded from zero.
  • Operators and existing accounts filtered out automatically before every send.

Next Case Study

Coupon Leak Detection & Dynamic Coupons

Found that 43% of a public return-trip coupon's discount was going to people it was never meant for — then replaced it with personal, single-use coupons issued automatically after each booking.