Ola Mabadeje from Cisco's Outshift group presents a product-focused overview of an AI-driven solution aimed at reducing failures in network change management. The core of this solution lies in a multi-agent system that interacts with a network knowledge graph and a digital twin of the production network. This system, designed with a natural language interface, allows network operations teams to manage changes by automating impact assessments, test plan generation, and test execution. By integrating with existing ITSM tools and leveraging an open-source collective for agent development, the goal is to create a scalable and flexible system that significantly reduces errors and improves efficiency in complex network environments. The presentation also highlights the importance of evaluation metrics for both the knowledge graph and agent performance to ensure tangible customer value.