How to Build Your Own Long-Term Agentic Memory System for LLMs - Mem0 from Scratch with DSPy

Learn to build a comprehensive long-term agentic memory system for Large Language Models from scratch using DSPy and QDrant Vector Database in this 52-minute tutorial. Discover how to recreate the core components of Mem0 by starting with a basic chatbot and progressively adding memory capabilities. Master the fundamentals of DSPy signatures and modules while implementing memory extraction from conversations, generating embeddings, and performing vector database operations with QDrant. Explore advanced concepts including tool calling with dspy.React, creating evaluation datasets, and building a memory upkeep CRUD agent for maintaining and updating stored memories. Follow along as the instructor demonstrates each component step-by-step, from initial setup through final testing, providing practical insights into building production-ready memory systems that can enhance LLM performance through persistent context retention.