A simple demo e-commerce application created from instructions on Claude using Next.js and Supabase. It uses Supabase vector to bring up related item when viewing a certain item.

Checkout the video walk through of creating this application using Claude: https://youtu.be/g5NjMqj22qI

This project uses:

• Claude to generate dummy data
• Next.js as the frontend
• Supabase database for storing item data and their embeddings
• Supabase storage to store item images
• OpenAI to generate images and embeddings

Generate dummy data

• Update the script/fashion_data.json file to update/ add dummy data
• Create a script/.env file and add OpenAI API key, Supabase URL, and Supabase service role key
• Run npm start inside the script directory to generate the images, embeddings, and store them in a Supabase project.

Run the application:

• Create a app/.env.local file and add your Supabase url and anon key
• Run npm run dev inside the app directory to run the app locally.

• Video walk through of the app
• Supabase AI & Vectors guide
• Supabase semantic search guide
• The missing pieces to your AI app (pgvector + RAG in prod) vidoe