FraudLens

FraudLens is an early scam detection and reporting platform built by our team for the PIDI Hackathon, where it advanced to the semifinals. I worked as the Frontend Developer, building the web application with ReactJS and connecting it to the team's backend services.

Online scams often spread faster than victims can react, and reporting them is usually confusing: people are unsure whether a message is really a scam, where to report it, and what happens after they do. FraudLens shortens that journey by meeting users where they already are, starting with a Telegram chatbot.

Users forward a suspicious message or link to the FraudLens bot, which analyzes it with Google Gemini through the AI SDK and replies with a scam probability score as a percentage, so users know how risky it is before they act. When the score is high, the bot gives the user a verified link to report the case directly on the FraudLens website, without having to start over.

On the web side, I built the reporting and tracking experience. The multi-step report form is pre-filled through the bot session: the reporter's phone number is verified automatically via Telegram, ID card details can be extracted with OCR, and the bot verification token keeps the report tied to the original conversation. After submitting, users receive a tracking code and can follow their report on a status page with a step-by-step progress tracker (Received, Analyst Verification, Investigation, Completed), an activity timeline, an incident summary with severity level, and a printable proof of report while they wait for it to be accepted.