An Architect's Guide to AI Agent Lifecycle Management and the SAP Family of Models

#AutonomousEnterpriseArchitect

As AI agents and MCP servers proliferate across enterprise landscapes, organisations face a defining architectural challenge: how do you discover, govern, secure, and measure the business impact of hundreds or thousands of autonomous agents? Without a unified approach, agent sprawl introduces unmanaged risk, architectural drift, and an inability to demonstrate ROI.
This session explores SAP's end-to-end approach to AI agent lifecycle management through the SAP AI Agent Hub (built on LeanIX). From discovery and registration through governance, observability, and decommissioning, you will see how SAP orchestrates its platform capabilities to register, observe, process-mine, assess, and map AI agents to enterprise architecture constructs linking agents directly to business capabilities, value streams, and organisational context. Whether you are building agents, securing them, or justifying their value to the business, this is how enterprise-grade AI agent governance works in practice.

After participating in this live session, you will be able to:

• Understand the full agent lifecycle (discover, register, govern, observe, decommission) and how it maps to your existing EA governance frameworks
• Learn how the SAP AI Agent Hub on LeanIX connects agents to business capabilities, value streams, and organisational context
• Learn what makes SAP's purpose-built models structurally different from generic LLMs and when to use them vs. third-party foundation models
• Identify the governance operating model your organisation needs before scaling from tens to hundreds of agents