Agentic AI with Chatops for Large Scale Network Operations

Operating large-scale networks necessitates interaction with numerous tools and dashboards for troubleshooting and analysis. Given the sensitivity of network services, minimizing manual interaction time is crucial for improving Mean Time To Repair (MTTR) and ensuring service availability. Recent developments in generative AI enable the transformation of interactions with diverse APIs and dashboards into intent-driven operations. Agentic AI offers capabilities to translate user intent into automated processes, thereby reducing the time engineers spend interacting with various interfaces and dashboards. This paper presents a novel approach for operating large-scale network infrastructure, wherein AI agents execute tasks involving API and dashboard interactions based on natural language commands. We evaluated the impact of this approach within a major global network environment. Our findings demonstrate significant time savings, consequently improving MTTR and overall service availability in this large-scale network.