Video by Hugging Face via YouTube
Quick case study persistent memory for AI agents featuring Mem0. The architecture of long form memory for agents goes far beyond a vector search. This deep dive explains Mem0’s memory stores, ingestion pipeline, hybrid retrieval with semantic search, BM25 and entity boosting, plus practical open models for running the system locally.
—
đź”— *Links*
– Mem0 documentation: https://docs.mem0.ai/
– Mem0 source code: https://github.com/mem0ai/mem0
– Hugging Face models: https://huggingface.co/models
– Hugging Face embedding models: https://huggingface.co/models?pipeline_tag=feature-extraction
– MTEB leaderboard: https://huggingface.co/spaces/mteb/leaderboard
—
đź‘‹ *Connect with me*
– My website: https://alejandro-ao.com/
– X (Twitter): https://x.com/_alejandroao
– LinkedIn: https://www.linkedin.com/in/alejandro-ao/
—
🤓 *Topics Covered*
– Long-term memory architecture for AI agents
– Mem0 ingestion, storage and hybrid retrieval
– Local LLM and embedding model choices
—
⏱️ *Timestamps*
0:00 Introduction
0:55 What agent memory is
3:28 Persistent memory across agents
4:26 Databases used by Mem0
8:06 Memory ingestion
15:30 Retrieval: queries and semantic search
19:19 Retrieval: BM25 and entity boosting
25:38 Open models and conclusion