Inspiration

Every B2B outreach team wastes money on lead lists that look big but barely convert — generic scraped contacts, wrong company size, wrong market, dead emails. We’d seen it firsthand doing outreach ourselves: you buy 500 leads, call them, and most are a wrong fit from the start. We wanted to flip the model — sell leads that are already validated against what you’re actually selling, so you only pay for ones that fit.

What it does

Buuzle finds the right leads for a company’s outreach, not just a lot of leads. A client enters their company URL, and Buuzle auto-fills an Ideal Customer Profile — industry, geography, job titles to target — pulled straight from their site. From there it builds a list of leads checked against that profile on firmographic fit, role & seniority, buying signals, verified work email, exclusions/DNC, and freshness. Leads that fail a check are dropped and shown why (too small, wrong market, personal email, competitor). The client only pays for leads that pass, delivered as a CSV, with an option to push straight into HubSpot, Pipedrive, or Salesforce.

How we built it

Next.js frontend and API routes for the ICP wizard, enrichment, and dashboard (orders, billing, files). The enrichment step calls out to pull firmographic data from a company’s domain and pre-fills the ICP form. Lead validation runs each candidate through the fit checks and scores a match percentage before it’s billed. We built the pitch video with a custom Python/ffmpeg motion-graphics pipeline layered over our own real screen-recorded product demo.

Challenges we ran into

Making the “pay only for what fits” promise actually visible in the product, not just a pricing line — we had to surface the validation reasoning (which check failed, why) rather than hide it. Tight hackathon timeline meant prioritizing the enrichment → validated-list flow over polish everywhere else.

Accomplishments that we're proud of

A working end-to-end flow — paste a domain, get an auto-filled ICP, get a real validated lead list with transparent pass/fail reasons and live pricing tied to accepted leads — not a mockup. Also the ad, which pairs a real demo of the actual product with motion graphics in the time we had.

What we learned

How much the “transparency on rejection” piece matters for trust in a lead-gen product — showing why a lead didn’t qualify sells the validation harder than the leads themselves do. Also a lot about ffmpeg filtergraphs and building explainer-style product video entirely from a script. Expand the fit-check signals (tech stack, intent/buying signals), add more CRM integrations, and move from manual enrichment review toward a fully self-serve, instant-checkout flow for smaller accounts.

What's next for Buuzle

Built With

  • claude
  • elevenlabs
  • firecrawler
  • github
Share this project:

Updates

Submission history