BobSec – Multi‑Agent Scam Shield

BobSec is an IBM Bob–powered, multi‑agent scam shield designed for everyday Indian users who are overwhelmed by scam calls, fake KYC SMS, fraudulent job offers and UPI payment traps. Instead of expecting people to understand cybersecurity jargon, BobSec lets them paste or forward any message, link, phone number or UPI detail into a simple web app. In under a few seconds, BobSec returns a verdict – Safe, Suspicious or Scam – and explains why in plain language, along with what to do next, such as not replying, blocking the sender or reporting to cybercrime authorities. Under the hood, IBM Bob acts as the orchestrator, routing each request to specialised agents such as ScamAgent for message and link analysis, FraudAgent for UPI and financial fraud patterns, IntelAgent for external threat intelligence, and PolicyCheckAgent for governance and safety rules. The backend is built in Node.js and Express, with a React frontend that focuses on accessibility and clear UX. The prototype includes demo‑mode safety, synthetic users, and preloaded Indian scam samples (KYC phishing and job scams) so judges can experience the full flow without real PII or credentials. BobSec emphasises governance and trust: every analysis is logged with an audit trail, the reasoning is transparent, and the system is designed to be hardened for production by re‑enabling strict authentication, rate limiting and live CTI and deepfake integrations. The result is a working IBM Bob multi‑agent system that is technically credible, aligned with the hackathon’s agentic theme, and grounded in the real pain of Indian users who lose money daily to sophisticated scams.