Sven-Erik Nielsen shows how to use GitHub Copilot and MCP servers to interrogate support tickets for themes and insights without pre-built dashboards. “As PMs we all know that support tickets are just an incredible source of insight, but sometimes it can be a little bit difficult to get them or the interfaces may not be conducive to really fluid analysis,” he explains. Sven uses MCP (Model Context Protocol) servers to connect to databases. “I already installed MCP servers beforehand. There’s lots of good literature out there on the Internet how to do that.” He starts them up and puts his query in the chat pane using agent mode, which is key for having the agent manipulate files and create changes to markdown. His prompt: “I want you to reason over this storm events table and help me understand the themes and insights. I don’t want you to do the work for me of determining what I should know and think about these incidents. Think of these as though they are support tickets. I need you to be kind of a thought partner here.” The LLM connects to the database through the MCP server and goes through a series of steps. “It’s kinda cool because you can look at basically the chain of thought. That’s a good way to double check you’re getting the right information, and you can use that to iterate and improve your prompt.” The output reveals themes: infrastructure vulnerability, tree damage, homes that are vulnerable, water hazards, meteorological patterns. “It gives me a sense of these incidents. It does give me some specific instances to go back in, like these event IDs here, to drill back into the support database.” “That way I don’t have to be beholden to a pre-built dashboard, which kind of constrains my thinking. On the other hand, sometimes, while I love writing SQL queries, it can be not the most fluid way of trying to think through what’s going on in the data. So it’s kind of a hybrid of those two approaches.” ➡️ Install MCP servers to connect your IDE to databases. Query support tickets conversationally instead of writing SQL. Get themes and specific ticket IDs to investigate without being constrained by pre-built dashboards.