Mydra logo
Artificial Intelligence
DeepLearning.AI logo

DeepLearning.AI

LLMs as Operating Systems: Agent Memory

  • up to 1 hour
  • Intermediate

This course teaches you how to build agents with long-term, persistent memory using Letta. Learn to manage and edit context efficiently, creating adaptive AI agents for real-world tasks. By the end, you'll have the tools to extend memory beyond the finite context window of LLMs.

  • Agent memory management
  • Letta framework
  • Multi-agent collaboration
  • Context window management
  • Persistent memory

Overview

In this course, you will learn to build agentic memory into your applications using the Letta framework. Discover how an LLM agent can act as an operating system to manage memory, autonomously optimizing context use. You'll explore the key ideas behind the MemGPT paper, including the two tiers of memory and how agent states are turned into prompts. By the end, you'll be equipped to create adaptive, collaborative AI agents for tasks like research and HR.

  • Web Streamline Icon: https://streamlinehq.com
    Online
    course location
  • Layers 1 Streamline Icon: https://streamlinehq.com
    English
    course language
  • Self-paced
    course format
  • Live classes
    delivered online

Who is this course for?

AI Enthusiasts

Individuals interested in learning how autonomous agents can manage their own memory.

Developers

Developers looking to build applications with agentic memory using LLMs.

Data Scientists

Data scientists who want to explore advanced memory management in AI applications.

Unlock the potential of LLMs by learning to manage agent memory effectively. This course covers key concepts like persistent memory and multi-agent collaboration, ideal for AI enthusiasts and developers. Enhance your skills and create adaptive AI agents for real-world applications.

Pre-Requisites

1 / 2

  • Basic Python skills

  • Interest in autonomous agents

What will you learn?

Introduction
Overview of the course and its objectives.
Editable Memory
Learn about editable memory with video and code examples.
Understanding MemGPT
Explore the key ideas behind the MemGPT paper.
Building Agents with Memory
Learn to build agents with memory using video and code examples.
Programming Agent Memory
Understand how to program agent memory with practical examples.
Agentic RAG and External Memory
Learn about agentic RAG and external memory with video and code examples.
Multi-agent Orchestration
Explore multi-agent orchestration with video and code examples.
Conclusion
Summarize the course learnings and outcomes.
Appendix - Tips, Help, and Download
Additional resources and code examples for further learning.

Meet your instructors

  • Charles Packer

    Co-founder & CEO, Letta

    Charles Packer is a co-founder and CEO at Letta. He is building the next generation of open AI systems at Letta.

  • Sarah Wooders

    Co-founder & CTO, Letta

    Sarah Wooders is the co-founder and CTO of Letta, a company that makes building real-world AI easy and accessible to businesses worldwide. She has deep expertise in product engineering with a passion for building products with cutting-edge research technologies.

Upcoming cohorts

  • Dates

    start now

Free