Embeddinghub logo

Embeddinghub

Stores embeddings durably, allows approximate nearest neighbor operations, enables partitioning, sub-indices, and averaging, and manages versioning, access control, and rollbacks.

Made by shabnam mokhtarani

  • Developer Tools

  • Tech

  • Artificial Intelligence

  • GitHub

What is Embeddinghub?

Embeddinghub is an open-source vector database designed to streamline the deployment of machine learning models in production environments. Built with four key objectives in mind, this database offers a robust solution for storing, managing, and leveraging machine learning embeddings

Highlights

  • Durable and highly available storage for machine learning embeddings
  • Efficient approximate nearest neighbor operations for rapid data retrieval
  • Advanced functionality, including partitioning, sub-indices, and averaging
  • Seamless management of versioning, access control, and rollbacks

Platforms

  • Web

Social

Features

    • Encode different business logic and user

    • Keep track of how embeddings are being used,

    • Run approximate nearest neighbor lookups, average

    • Store and index billions vectors embeddings from

    • Create, manage, and rollback different versions