Privacy Shield

Project Dashboard

Every website we visit deploys dozens of invisible trackers, analytics scripts, advertising pixels, social media beacons — silently profiling our behavior across the internet. Most people have no idea how many companies are watching them, or which websites expose them the most. We wanted to flip the script: give users full visibility into who's tracking them, the power to block it, and an AI-driven security score to understand their real risk level.

What it does Privacy Shield is a Chrome extension + AI-powered dashboard that detects, blocks, and analyzes web trackers in real time.

Real-time tracker detection — Monitors every network request, matching against thousands of known trackers from DuckDuckGo's Tracker Radar database plus custom pattern matching for tracking endpoints (/pixel, /beacon, /collect, etc.) One-click blocking — Users manage a personal blocklist that selectively blocks third-party tracker requests across all sites Analytics dashboard — A four-page React dashboard shows an overview of tracking activity, breakdowns by company/category/domain, historical trends (7d / 30d / 6m), and a searchable tracker table AI Security Report — Google Gemini analyzes your browsing data and generates a 0–100 security score per domain, highlights your three riskiest sites, and provides a plain-English summary of your privacy posture Per-user data isolation — All data is scoped by email via Auth0 authentication, so every user sees only their own tracking history How we built it Browser Extension: Chrome Manifest V3 with a background service worker that intercepts network requests, detects trackers via a hybrid local + DuckDuckGo database approach, and syncs data every 20 seconds Frontend: React 18 + TypeScript + Vite + Tailwind CSS, with Recharts for data visualization (line charts, bar charts, ring gauges) Backend: FastAPI (Python) serving REST endpoints for tracker ingestion, blocklist management, breakdown analytics, and AI-powered security reports Database: Snowflake cloud data warehouse for scalable, email-scoped storage of all tracker and cookie events AI Analysis: Google Gemini Flash Lite scores domains based on tracker category, hit frequency, and cross-site spread, returning structured JSON with per-domain risk assessments Auth: Auth0 OAuth2 for secure, passwordless authentication What's next for Privacy Shield Cookie deep-analysis — Enable full cookie tracking with elevated permissions to show cookie lifetimes, third-party cookie chains, and fingerprinting attempts ML-based anomaly detection — Flag unusual spikes in tracking activity that could indicate compromised sites or new tracker deployments Cross-user anonymized insights — Aggregate anonymized data across users to build a community-driven tracker reputation system Firefox & Safari support — Port the extension to other browsers using WebExtensions API Export & compliance tools — Let users export their tracking reports for GDPR/CCPA data subject access requests Mobile companion app — Extend privacy monitoring to mobile browsing with a lightweight companion app. We faced several challenges when trying to complete this program. It was hard to coordinate with students remotely from different universities, a lot of time was spent successfully collaborating.While trying to integrate everyones code together, we spent a lot of time debugging errors, on each of our platforms. Finally, setting up the backend to interact with the database took a lot of new learning.