Storecraft + MongoDB Livestream of Semantic Vector Search extension and frontend

---

This repo was created during the livestream we had th the official MongoDB YouTube channel.

We demo the usage of Storecraft with Mongo Vector Search

What we did:

• We created an extension for Storecraft, that adds semantic/ai search for products catalog.
• We implemented a small frontend react application to demonstrate the vector search.

How to run

First,

First, make sure you have the following environment variables

//.env

OPENAI="YOUR OPENAI KEY"
MONGODB_URL="YOUR MONGODB CONNECTION URL"

Then,

npm install
npm run migrate
npm start

Setup semantic/ai vector index in Mongo Atlas

1. in Atlas dashboard, create a vector index (call it vector_index) for products collection:

{
  "fields": [
    {
      "numDimensions": 1536,
      "path": "embedding",
      "similarity": "cosine",
      "type": "vector"
    }
  ]
}

Now,

• Open dashboard at http://localhost:8080/api/dashboard
• Upsert some products so every upserted product will be indexed for semantic search by it's title and description.

Webapp

And then, follow the instructions in the ./frontend folder to start the web app.