Building memory for AI Agents
summary
Cognee is a Python SDK that connects LLMs to graph and vector stores and claims to solve memory consistency at scale by combining Redis short‑term cache, persisted graph skeletons and embedding indexes. It runs modular ingestion pipelines, entity extraction, ontology merging and a self‑improvement layer that distills session data into long‑term skeletons and applies retention rules. Demo uses Langraph tools; deployed usage hits ~286k monthly memory operations now trending toward 500k across ~63 companies, and the system is production‑ready.
AI-written summary · about
chapters (5)
- 0:00Introduction and background
- 3:42Building Cogni and AI memory
- 8:58Cogni architecture and self-improvement
- 11:48Demo with Langraph and memory integration
- 17:32Q&A and closing
related talks
26mStore Your AI Agent's Memory and Context in a Knowledge Graph
51mMemory Masterclass: Make Your AI Agents Remember What They Do!
25mWhat does it take to build long-term state for AI agents?
29mThe AI Agent Memory Landscape
46mThe Anatomy of Memory in Humans and AI Agents
14mOptimizing RAG with Semantic Caching & LLM Memory
27mMemory Engineering: Going Beyond Context Engineering
52mContext Engineering with Redis and LangChain