Cognee builds persistent memory for AI agents. Its ECL pipeline (Extract, Cognify, Load) ingests data from more than 38 sources and structures it into a knowledge graph with embeddings and relationships that can then be searched, rather than relying on a bare vector store. A 'memify' layer refines the graph through feedback loops, feeding rated responses back into edge weights so retrieval sharpens with use. Cognee unifies relational, vector and graph storage into a single engine and plugs into agent frameworks including the Claude Agent SDK, OpenAI Agents SDK, LangGraph, Google ADK and n8n, as well as Amazon Neptune and Neo4j. The company reported that pipeline volume grew from roughly 2,000 runs to over one million during 2025, that Cognee runs in more than 70 companies, and that the open-source project has over 12,000 GitHub stars and 80+ contributors. Named users include Bayer, the University of Wyoming and dltHub. Version 1.0, announced in June 2026, introduced a memory-native API built around remember, recall, improve and forget, and the ability to run the full memory layer on a single Postgres instance.
Funding
Cognee announced a $7.5 million seed round on February 19, 2026, led by Pebblebed, with participation from 42CAP and Vermilion Ventures plus angel investors from Google DeepMind, n8n and Snowplow.