PolicyPulse AI — Autonomous Compliance Monitoring Platform
Link to open source: https://github.com/Md-javid/HACKFEST-2-0
PolicyPulse AI is an autonomous compliance monitoring platform that eliminates the manual burden of regulatory compliance for organizations.
The Problem: Companies spend thousands of hours manually reviewing compliance policies, cross-checking employee records, vendor contracts, and infrastructure configurations against GDPR, SOC 2, and ISO 27001 requirements. Violations go undetected for months, resulting in heavy fines and data breaches.
What We Built: PolicyPulse AI lets you upload any compliance policy PDF — it automatically extracts enforceable rules using Google Gemini AI, then continuously scans your company database for violations with confidence scoring and severity classification.
Key Features:
- AI Policy Ingestion — Upload a PDF and Gemini extracts structured compliance rules automatically
- Autonomous Violation Detection — Rule engine scans all company records (employees, servers, vendors, data stores) and flags violations in real time
- Multi-Agent ReAct System — 4 specialist AI agents (Security, Privacy, Vendor, Operations) autonomously remediate violations using a perceive → reason → act → reflect loop
- Proactive Risk Prediction — Scans records before a violation occurs and predicts upcoming compliance risks
- AI Policy Advisor — Analyzes violation patterns and suggests new policy rules to close coverage gaps
- Compliance Dashboard — Live compliance score, violation trends, audit reports with Gemini-powered executive summaries
- Human Oversight — Full audit trail; every AI action can be reviewed, approved, or escalated
Tech Stack: React 19 + TypeScript + Vite (frontend), FastAPI + Python (backend), MongoDB, Google Gemini 1.5 Flash, JWT auth with role-based access control.
Built for organizations that need compliance to be continuous, automated, and explainable — not a quarterly checkbox exercise.
This build was uploaded as a hackathon project
