Obtenez 3 mois à 0,99 $/mois

OFFRE D'UNE DURÉE LIMITÉE
Page de couverture de Agentic Memory: Stateful AI & RAG Extensions (Chapter 16)

Agentic Memory: Stateful AI & RAG Extensions (Chapter 16)

Agentic Memory: Stateful AI & RAG Extensions (Chapter 16)

Écouter gratuitement

Voir les détails du balado

À propos de cet audio

Discover how agentic memory is transforming AI from forgetful assistants into adaptive, stateful partners that remember, learn, and evolve over time. In this episode, we unpack Chapter 16 of Keith Bourne’s 'Unlocking Data with Generative AI and RAG' and explore the strategic impact of extending Retrieval-Augmented Generation (RAG) with dynamic memory systems designed for real-world business advantage.

In this episode:

- What agentic memory is and why it matters for AI-driven products and services

- Comparison of leading agentic memory tools: Mem0, LangMem, Zep, and Graphiti

- How different memory types (working, episodic, semantic, procedural) enable smarter AI agents

- Real-world use cases across finance, healthcare, education, and tech support

- Technical architecture insights and key trade-offs for leadership decisions

- Challenges around memory maintenance, privacy, and compliance


Key tools & technologies mentioned:

- Mem0

- LangMem

- Zep

- Graphiti

- Vector databases

- Knowledge graphs


Timestamps:

0:00 - Introduction to Agentic Memory & RAG

3:30 - The strategic shift: from forgetful bots to adaptive AI partners

6:00 - Why now? Advances enabling stateful AI

8:30 - The CoALA framework: modeling AI memory like human cognition

11:00 - Tool head-to-head: Mem0, LangMem, Zep/Graphiti

14:00 - Under the hood: memory extraction and storage techniques

16:00 - Business impact: accuracy, latency, ROI

17:30 - Reality check: challenges and risks

19:00 - Real-world applications & leadership takeaways


Resources:

- "Unlocking Data with Generative AI and RAG" by Keith Bourne - Search for 'Keith Bourne' on Amazon and grab the 2nd edition

- Memriq AI - https://memriq.ai

Pas encore de commentaire