squaredev-io / embeddings-engine

Create, store and enable embeddings related features like semantic search, RAG, recommendation, image to image and more

Embeddings engine

Squaredev embeddings engine is an API first product that simplifies the process of creating and storing embeddings. It also, enables features like semantic search, recommendations, RAG, image-to-image search, etc through a simple API and SQK.

Embedding engine is based in Python and Postgres making it reliable, scalable and easy to deploy.

Features

Text

• 🟦 Embeddings creation and storage through open source state of the art models.
• 🟦 Semantic search
• 🟦 Recommendations
• 🟦 Retrieval Augmented Generation (RAG)
• 🟦 Clustering

coming soon

• 🟦 Images / multimodal
• 🟦 Image-to-image search
• 🟦 Image-to-text search
• 🟦 Image-to-image recommendations

Getting started

We are still under development, but you can try our API by following the instructions below.

1. Clone the project

git clone https://github.com/squaredev-io/embeddings-engine.git
cd embeddings-engine

2. Create a virtual environment adn activate it

python3 -m venv .venv
source .venv/bin/activate

3. Install the dependencies

make install

4. Run the database

make db

5. Run the server

make run

6. Open your browser and go to http://localhost:8000/docs