Back to Projects
An AI honeypot API that detects scam patterns, extracts useful intelligence, and keeps fraudulent conversations engaged safely.
Honey Comb combines deterministic entity extraction with an LLM-driven conversation agent. It recognizes common Indian scam patterns, identifies red flags, extracts phone numbers, UPI IDs, links, and case references, and maintains a natural multi-turn response strategy.
Core Features
Scam classification across phishing, UPI, job, lottery, and impersonation fraud
Structured extraction of financial identifiers, contact details, and malicious links
Context-aware multi-turn engagement powered by Groq
Red-flag analysis and final session intelligence summary
Serverless API designed for fast integration with messaging channels
Built With
Node.js
Express
Groq
Llama
Security
AIConversation agent
APIIntegration
Multi-turnContext
Top 2%Summit selection
Project Previews
Illustrative interface concepts for Honey Comb.
Preview 2 of 3