HN Showcase
Back to browse
Verak – Fake Seller Detection for Digital Marketplaces

Verak – Fake Seller Detection for Digital Marketplaces

by Mikewillcodes·Mar 6, 2026·1 point·0 comments
Visit ProjectView on HN

AI Analysis

●●SolidBig BrainNiche Gem

Clusters coordinated fraud rings via timing, infrastructure, referral patterns—real forensic thinking.

Strengths
  • •Behavioral clustering approach is genuinely clever: creation timing, email patterns, shared infrastructure, referral loops are non-obvious fraud signals.
  • •Interactive demo on real data (5,200 accounts, 7 clusters) lets you poke the engine without signup—transparency builds credibility.
Weaknesses
  • •No public pricing or deployed customers visible; reads as early-stage enterprise product, not ready-to-buy.
  • •Marketplace fraud detection is a known problem (Stripe Radar, Sift, Riskified); positioning unclear against established vendors.
Category
Security
Target Audience

Marketplace platforms, ecommerce operators, fraud teams at mid-size consumer/B2B marketplaces.

Similar To

Stripe Radar · Sift Science · Riskified

Similar Projects

AI/ML●●Solid

Facebook Marketplace MCP

Uses internal GraphQL protocol directly instead of slow browser automation.

Big BrainNiche Gem
rpst
121mo ago