Looksy

What We Do

AI-powered product discovery platform that surfaces trending products through personalized, swipe-based recommendations using machine learning.

Business Model

Discovery platform powered by AI/ML—we don't process transactions. Users discover products on Looksy, then purchase directly from retailers.

Problem We Solve

Finding trending products is fragmented across hundreds of sites. Looksy uses ML to aggregate, analyze, and personalize product discovery in one addictive feed.

Target Audience

Fashion-conscious millennials and Gen Z who prefer swipe-based discovery over traditional search.

How It Works

1. 1. AI curates trending products
2. 2. Swipe through personalized feed
3. 3. Save products to collections
4. 4. Shop externally on retailers' sites

ML Stack

• •ByteDance Monolith framework for recommendation engine
• •OpenAI CLIP for product parsing with vision AI
• •Marqo embeddings for semantic search

Infrastructure

• •Google Cloud Run infrastructure
• •Automated web scraping systems
• •Scalable data processing pipelines

Platform

• •Mobile-first web application
• •iOS app (launching Jan 2026)
• •Thousands of products indexed

Product Stage

Pre-launch Beta with fully functional web app and ML recommendation system. iOS app launching January 2026.

Key Achievements

• ✓Built entire ML recommendation engine using ByteDance's Monolith framework
• ✓Developed web scraping infrastructure and automated product parsing systems
• ✓Architected cloud-based data processing pipelines

Background

• •Technical product development
• •Retail industry experience
• •Based in United Kingdom