Belgrade MUG #9: Memory for AI Applications (Skill Badge Session)

MongoDB User Group Belgrade is once again taking part in Belgrade AI Week this year, organized by Startit!

As part of this year’s program, we are hosting a technical session focused on one of the most important building blocks of modern AI systems: memory.

AI agents are only as useful as their ability to remember.

At the end of the event, attendees will have the opportunity to complete the official MongoDB Skill Badge assessment and earn their certification.

No laptops or local setup are required for participation, skill test can be taken using mobile phones.

We will cover the core concepts behind AI memory systems, including conversational memory, semantic caching, checkpointing, and long-term memory retrieval. We will also discuss how memory transforms stateless LLM applications into production-ready AI agents.

The session will explore architectures and technologies such as:
• MongoDB as the persistent memory and vector storage layer
• LangGraph for orchestration and checkpointing
• MongoDB Vector Search for long-term semantic memory
• Modern LLMs such as Claude, Gemini, or GPT as reasoning engines

⏱ Agenda
|18:00 PM - 18:15 PM| - Gathering, networking, food
|18:15 PM - 19:00 PM| - Memory for AI Applications Session
|19:00 PM - 19:15 PM| - Skill Badge Test
|19:15 PM - 20:00 PM| - Networking, food, swags...

🧠 What you will learn
During this session you will:
• Understand short-term and long-term memory in AI applications
• Learn how checkpointing and semantic caching work
• Understand how conversational memory is implemented in AI systems
• Explore long-term memory using MongoDB Vector Search
• Learn how to design production-grade memory architectures for AI systems