AI Native ODA Canvas скачать в хорошем качестве

AI Native ODA Canvas

Discover how the Open Digital Architecture (ODA) is evolving to support AI-native telecommunications systems at enterprise scale. This video explores the Innovation Hub's work on building an AI-native canvas that enables: • Exposing ODA components as tools for AI agents using MCP (Model Context Protocol) • Deploying AI agents themselves as modular, composable components • Operating multi-vendor, multi-agent architectures with governance and control Learn about the architectural patterns emerging from TM Forum Catalyst programs, including how to extend the component model for agentic components, implement AI gateways for responsible AI practices, and establish continuous evaluation frameworks for non-deterministic AI systems. Key topics covered: MCP integration with TM Forum Open APIs Agentic component architecture AI-native canvas services Multi-agent orchestration at scale Classical ML and GenAI model management Evaluation frameworks and responsible AI governance Perfect for telecommunications architects, AI engineers, and technology leaders interested in operationalizing AI agents within enterprise telecommunications environments. Chapters: 0:00 Introduction 1:58 Lessons learned 5:30 MCP support 9:17 Agents packaged as composable ODA Components 15:53 Canvas services to support AI Agents at scale 17:21 End-to-end Data and AI Architecture 22:10 Canvas Resource Inventory 24:54 Classical ML Models 25:56 Evaluation and testing