Skip to Content
View all events

AI Memory Management in Agentic Systems

Published by O'Reilly Media, Inc.

Intermediate content levelIntermediate

Empowering AI agents with robust memory

Course outcomes

  • Develop a robust mental model of AI memory management by understanding and differentiating between short-term, long-term, episodic, semantic, and procedural memory components
  • Write code to initialize, segment, structure, retrieve, update, delete, and create memory within AI agents using state-of-the-art techniques
  • Utilize retrieval-augmented generation (RAG) techniques—including lexical, vector, and hybrid retrieval approaches—to efficiently integrate memory into LLM-based applications
  • Configure and implement Oracle AI Database as a scalable and flexible memory provider, incorporating best practices in document storage, indexing, and vector search
  • Build end-to-end AI agents that use memory as a core resource for decision-making, planning, and long-term adaptation, drawing on real-world case studies and recent research
  • Analyze trade-offs between different memory architectures and choose appropriate strategies for specific application scenarios

Course description

Join expert Richmond Alake to understand the theoretical foundations, state-of-the-art research, and practical implementations of memory management in AI. Starting with a conceptual understanding of what memory means for an AI agent and how it parallels human cognition, you’ll then explore the core components that constitute robust memory management systems—ranging from initialization and segmentation to retrieval, updating, deletion, and creation. You’ll also examine advanced retrieval-augmented generation techniques—including lexical, vector-based, and hybrid approaches—as well as the emerging concept of agentic RAG. Finally, you’ll gain hands-on insights into how databases, and Oracle AI Database in particular, serve as powerful memory providers and enable scalable, efficient, and secure long-term memory architectures in AI applications.

What you’ll learn and how you can apply it

  • Learn what constitutes memory in AI, including key cognitive components and analogies with human memory
  • Understand mechanisms of initialization, segmentation, structure, retrieval, update, deletion, and creation as they apply to AI agents
  • Review cutting-edge approaches and frameworks in memory management such as generative agents, A-Mem, MemGPT, and agent workflow memory
  • Learn the theory and practice behind retrieval-augmented generation methods, including hands-on examples of lexical, vector, and hybrid retrieval approaches
  • Explore how to use Oracle AI Database as the backbone for persistent and scalable memory storage, covering document modeling, indexing strategies, and native vector search capabilities
  • Understand how to embed memory management components into agent workflows for real-time decision-making and long-horizon planning
  • Build more context-aware and adaptive AI systems
  • Enhance existing LLM-based products with persistent and dynamic memory
  • Design systems for long-term learning and personalization
  • Optimize memory retrieval and update strategies to reduce computational overhead and improve agent realism

This live event is for you because...

  • You’re an AI or ML engineer who works on LLM-based applications and needs to implement persistent memory for complex tasks.
  • You’re a data scientist who builds intelligent systems and wants to understand how to enhance models with robust memory management.
  • You’re a software developer in the AI space who needs to implement scalable and efficient memory solutions for agent-based applications.
  • You’re an experienced AI practitioner looking to advance your knowledge of state-of-the-art memory architectures for agentic systems.

Prerequisites

  • Familiarity with large language models (e.g., GPT-3.5, GPT-4) and their limitations (context windows, tokenization)
  • Proficiency in Python, along with experience using common AI/ML libraries
  • Basic experience with databases (preferably NoSQL databases like Oracle AI Database) and data storage concepts
  • A general understanding of concepts like embeddings, vector search, and retrieval-augmented generation (RAG) will be helpful
  • (Optional) Exposure to cognitive architectures (e.g., Soar, ACT-R) or previous experience with AI agents can provide additional context but is not required

Recommended preparation:

  • Install Python (version 3.8 or higher)
  • Ensure you have access to a Oracle AI Database instance (either local or via Oracle AI Database Atlas) and familiarize yourself with its basic operations

Recommended follow-up:

Schedule

The time frames are only estimates and may vary according to how the class is progressing.

AI memory (55 minutes)

  • Presentation: Overview of AI memory concepts—short-term (context window) versus long-term (conversational memory), episodic versus semantic, and their differences; analogies to human cognition and memory; challenges of memory in AI agents and agentic systems; real-world examples and intuition behind AI memory
  • Group discussion: Which aspect of memory do you find most challenging to conceptualize?
  • Q&A
  • Break

Memory management (65 minutes)

  • Presentation: Overview of retrieval-augmented generation (RAG) methods—lexical, vector, hybrid, agentic RAG, and rerankers; core components of memory management—initialization, segmentation, structure, creation, retrieval, update, deletion; best practices and design patterns in agentic systems; overview of memory design architectures from generative agents, A-Mem, MemGPT, and agent workflow memory
  • Group discussion: Which memory component do you believe is most critical for long-term agent coherence?
  • Q&A
  • Break

AI retrieval (60 minutes)

  • Presentation: Overview of AI retrieval—definition, concepts, and techniques; introduction to the concept of “memory providers” and their role in AI application architecture; examples of memory providers
  • Hands-on exercises: Setting up the environment; indexing and vector search implementation; building retrieval functions; end-to-end agentic RAG system; performance monitoring and logging implementation
  • Q&A
  • Break

Memory management technique implementation (50 minutes)

  • Hands-on exercises: Implementing memory management components using Oracle AI Database, LangGraph, and LangMem

Wrap-up and Q&A (10 minutes)

Your Instructor

Richmond Alake

Richmond Alake is director of AI developer experience at Oracle, where he leads developer relations for the Oracle AI Database and pioneers the discipline of memory engineering, a term he coined to describe the structured design of agent memory systems. He runs the #100DaysOfAgentMemory LinkedIn series, which has reached tens of thousands of AI practitioners.

Richmond has published over 200 technical articles with more than one million views across Medium, NVIDIA’s technical blog, and other industry publications, covering topics from agent memory systems to production LLM application development. He has spoken at the AI Engineer World’s Fair, DeveloperWeek, Confluence, and Monki Gras. Previously, he served as an AI/ML developer advocate at MongoDB, where he built MemoRizz, an open source Python agent memory library, and codeveloped a DeepLearning.AI course on prompt compression and query optimization. Richmond holds a master’s degree in computer vision, machine learning, and robotics from the University of Surrey.

linkedinXlinksearch

Skill covered

Generative AI