Building the unified storage layer for AI Agents and applications
Depths AI is building a unified storage layer called gyaan for AI Agents and applications. The aim of the package is to enable developers to integrate vector-embedding database, chat history, analytics logs, knowledge graphs and a memory agent layer on top it, within few lines of code.
Currently, gyaan is v0.1: The world’s cheapest S3-compatible vector embedding database.
The project is and will proudly remain open-source! For ease of usage, we encourage you to try out the free tier of the cloud version (currently in open-beta).
Leveraging a unified storage core for S3-compatible storage, we would introduce the remaining aspects of storage layer as:
We recommend using uv for managing your python projects. Command to add gyaan in an existing uv project is as follows
Or if you prefer pip installation for your python projects: