Catalyst AI Assistant & Automation with MCP Server
Solution overview
This lab explores the AI capabilities built into Cisco Catalyst Center through two complementary approaches. First, you'll get hands-on with the native AI Assistant, a conversational interface built directly into the Catalyst Center UI that lets you troubleshoot, monitor, and query your network using plain language instead of navigating dashboards or memorizing CLI commands. Working within a pre-configured SDA Fabric environment at site ATC56, you'll practice asking real troubleshooting and monitoring questions, comparing Default and Deep Reasoning modes, and validating the assistant's responses against actual device configuration and CLI output.
From there, the lab extends beyond the built-in UI to show how Catalyst Center's data and APIs can be reached externally through an MCP (Model Context Protocol) server, letting an outside AI client discover and call Catalyst Center's APIs directly rather than requiring a custom integration for each one. By the end of the lab, you'll have a practical sense of both paths: when to rely on the native AI Assistant for work that starts inside Catalyst Center, and when to use MCP for workflows that need to reach into Catalyst Center from the outside.