Teaching AI to Remember: Persistent LLM Memory with PostgreSQL & LangChain4J скачать в хорошем качестве

Teaching AI to Remember: Persistent LLM Memory with PostgreSQL & LangChain4J

Part 8 – Persisting LLM Memory with PostgreSQL (Java + LangChain4J) This session is all about giving our AI a memory! After the audience vote, we’re diving into persisting LLM memory in PostgreSQL so our AI-powered adventure games can remember past sessions — making storytelling and tool usage smarter than ever. 💡 If you’re a Java developer curious about Ollama + LangChain4J + persistent AI memory, this stream will show practical ways to store and retrieve AI state between sessions. 🎯 Key topics in Part 8: PostgreSQL setup & integration with LangChain4J Persisting & retrieving LLM memory Memory-aware prompt engineering for storytelling & tools Session handling improvements Lessons from memory-enabled AI adventures 🛠️ Tech Stack: Java 21+ · LangChain4J · Ollama (custom modelfiles) · PostgreSQL · LLaMA 3.2 · Vaadin · IntelliJ AI Assistant 📦 GitHub repo: 👉 https://github.com/JohannesRabauer/ai... 🎥 Subscribe + 🔔 for live Java AI coding: persistent memory, custom Ollama models, and AI-powered games. 00:00 Intro and Forecast 01:09 Preparations for PostgreSQL 04:49 Setting up the Data Model 12:40 Implementing the LangChain4J Memory Persistency 31:10 What about the Memory ID? 59:00 Keeping the Memory ID in the game 01:12:50 Why is the memory ID an object?! (Solution in the comments) 01:17:23 Starting the application (with interesting results) 01:24:15 Victory once again! 01:24:50 Final result and conclusion #Java #Ollama #LangChain4J #AIModels #LLMMemory #PostgreSQL #Qwen3 #LLaMA #AIGames #PersistentAI