HEYI.

HEYI (“Hey, AI!”) is a full-stack Chrome extension that helps shoppers spot AI-generated text and images on ecommerce pages—pairing a fine-tuned model with a MongoDB-backed warning system and proactive alerts.

Timeline: Academic collaboration — model training, extension, and backend
Role: Co-creator — ML pipeline, extension, API, and product narrative
Stack: Chrome Extension API · Node · Express · MongoDB · Hugging Face · Gemini

Get HEYI on the Chrome Web Store

Brand mark for the extension and onboarding.

Overview

Why it exists

AI-generated listings and images are increasingly hard to distinguish from real products. HEYI scans on-page text and images for AI signatures, with dual-mode analysis and proactive alerts when you visit domains with a history of generated content.

The model side uses a custom classifier trained on a large sample set, hosted for inference with Hugging Face; image workflows also lean on Google Gemini for complementary signal. The Express server routes inference and persists domain intelligence in MongoDB Atlas.

Extension

How it behaves in the browser

DOM analysis

Scripts scrape and evaluate page content, then call the backend for classification results.

Background service

Notifications and domain history connect users to warnings without blocking normal browsing.

Transparency-first

Clear messaging so shoppers can decide how much to trust a listing.

OpenTomorrow

Shared time-capsule product—Expo, Next.js, and Supabase for capture, social, and marketing.

All projects