Build a memory-enabled AI agent that remembers user preferences and learns from conversations using Redis and LangGraph.
The goal of this tutorial is to empower you with a horizontal concept that you can apply to your own agent use cases.
- Dual-Memory Architecture: Implement short-term (conversation state) and long-term (persistent knowledge) memory
- Semantic Search: Use RedisVL for semantic memory retrieval with embeddings
- Memory Types: Understand differences between episodic (user experiences) vs semantic (general knowledge) memory patterns
- Production Patterns: Tool-based memory management and conversation summarization
- LangGraph Integration: Build complete workflows with Redis checkpointers for state persistence
Agent Memory Tutorial Notebook
- OpenAI API Key (with billing enabled)
- Redis optionally installed in Colab, or use Redis Cloud
A travel agent that:
- Remembers user preferences across conversations
- Stores long term memories ("I prefer Delta airlines")
- Provides personalized recommendations based on past interactions
- Manages conversation context automatically
Total Tutorial Time: ~30-45 minutes
Difficulty: Intermediate (Python, LangGraph, Tool calling, other basic AI concepts)