VFlash: Fast, Accurate Visual Planning via Flash Primitive Decoding
u_Unrolled
All projects

Agents · Intermediate · by Mem0 AI

Mem0: Self-improving memory layer for AI agents

A persistent, self-improving memory infrastructure for LLMs and agents — dynamically extracting facts, profiling users, and updating memory over time.

1–2 weekends 69000 on GitHubPythonQdrantNeo4jOpenAI / Ollama
Open the code on GitHub

What you'll learn

  • Dynamic fact extraction and entity profiling from conversations
  • Layered memory: user, session, and agent-level stores
  • Memory conflict resolution and temporal decay strategies
  • Graph-backed relational memory with Neo4j integration

Read first

The lessons that explain what this project uses.

Part of the Agentic AI path.

Milestones

0/4

Built it? Show it.

Add your own repo to earn 150 XP and the Shipper badge. It appears on your profile, so it's ready for your resume.

More agents projects