The Future of AI Agents: How Standards and Evaluation Drive Innovation скачать в хорошем качестве

The Future of AI Agents: How Standards and Evaluation Drive Innovation

Multi-agent systems require seamless agent communication and robust evaluation. Join Galileo's ‪@erinmikail‬ and ‪@Cisco‬ to explore the infrastructure needed to build reliable, interoperable multi-agent systems, including an open, standardized framework for agent-to-agent collaboration. Discover how the AGNTCY collective (Galileo, ‪@LangChain‬, ‪@LlamaIndex‬, ‪@MongoDB‬, ‪@Weaviate‬, and more) is building the Internet of Agents, illustrated through a practical agent case study. 📢Watch our webinar to learn 🧠 • How to build interoperable AI agents across different frameworks • Standards-based approaches for agent communication • Robust agent evaluation techniques and performance monitoring 🌐 Help us build the Internet of Agents: https://agntcy.org/ 💻 Learn more about Galileo: www.galileo.ai 📩 Have questions? Reach out at [email protected] Chapters: 00:00 Introduction and Speaker Introductions 00:47 Housekeeping and Q&A Guidelines 01:32 Introduction to AGNTCY and Agent Communication 02:46 The Need for Standards in Agent Interoperability 04:24 Challenges and Solutions in Agent Communication 05:12 Cataloging and Describing Agents 07:04 Standards and Protocols for Agent Communication 17:45 Live Coding: Building an AGNTCY-Compatible Agent 25:15 Galileo's Role in Agent Evaluation 26:45 Prescriptive vs. Collaborative Agent Workflows 27:06 Observability and Security in Agent Interactions 27:42 Current Capabilities of Galileo 28:11 Evaluating Agent Performance 29:29 Live Demonstration of Galileo Integration 30:22 Real-Time Agent Interaction Example 33:43 Challenges and Solutions in Agent Collaboration 36:27 Final Thoughts and Takeaways 38:28 Q&A: Process Flow Tools vs. Rules-Based Approaches