Skip to Content
Introduction

Organise your data and get better responses from AI

Your data is like a haystack. The more you have, the harder it is to use it effectively.
Letting LLMs access any data you have in your system is inefficient, unsafe and not very flexible.
Cognee connects your data points and helps LLMs get the right context for the answer you need Improve the accuracy of your AI agents with Cognee. Connect the dots with modular tool that can ingest and manage structured, semi-structured and unstructured data and include human inspired memory algorithms

Cognee general demonstration

🏗️ Cognee Ecosystem

Cognee consists of two main repositories:

Cognee Core 

The main repository containing:

  • Core cognee functionality and APIs
  • Essential vector and graph database adapters (LanceDB, Neo4j, Kuzu)
  • Primary documentation and tutorials
  • Stable, production-ready features

Cognee Community 

The community-driven repository for:

  • Additional database adapters - Vector stores (Qdrant, Azure AI Search, Redis) and graph databases beyond the core ones
  • Experimental features - Cutting-edge functionality and community contributions

📚 Documentation Guide

  • Jump on a Quickstart to get cognee running

  • Follow a few Tutorials to learn about our UI, SDK and scaling cognee

  • How-to Guides provide you more patterns for advanced usage from technical and functional perspective

  • Core Concepts can help you gain a deeper understanding on the inner workings of the system

  • Use Cases show the business value you can gain from cognee

  • Reference will get you covered with technical details

  • Browse Research if you want to understand the academic background

  • Resources will help you from a beginner, an intermediate and an advanced perspective

  • Visit FAQ short bite-size information


Quickstart