Build an AI agent that joins every user research conversation and becomes a searchable institutional memory
Flo Crivello from Lindy shows how to build a “Lindy brain” - an agent that joins every user research conversation and becomes a resident expert on your customers. Imagine having one person who’s been in thousands of user research chats and can answer any question in 5 seconds about everything they’ve seen.Here’s Flo’s workflow for building this UXR agent:First, set up a calendar trigger. When a calendar event starts containing “UXR” in the name, Lindy automatically joins and records the meeting. Flo emphasizes: “I’m going to do nothing - it’s just going to automatically figure out what to join.”Second, configure post-meeting actions with minimal setup. After recording, the agent appends a row to a spreadsheet. The magic: “I’m going to configure none of this. It’s just going to be able to figure out what I want to add based on the header of the spreadsheet.” If you add a new column like “satisfaction on a scale of one to 10,” Lindy automatically populates it without changing configuration.Third, create meeting notes automatically. Lindy creates a Google Doc in your UXR notes folder with a simple prompt: “Enter the summary notes for this user research call.”Fourth, enable RAG (retrieval augmented generation) to search across all conversations. Add a chat trigger so you can ask questions about your entire research corpus. Flo demonstrates asking: “What did people say about their needs on the compliance side?” and getting specific answers from past conversations.Fifth, integrate with Slack for team access. Set up a filter for “@UXR Lindy” mentions, then the agent searches the knowledge base and replies in thread. Your whole team can query customer insights directly from Slack: “Lindy brain, what do you think?”Bonus automation: Set a weekly trigger for Friday at 5 PM to send a summary of all user research notes from that week.Flo shows a real example from their own Lindy brain: when they released email attachments, they asked which customers wanted this feature and got a detailed list with names, dates, and use cases for each customer who mentioned it.➡️ Stop relying on scattered meeting notes. Build an AI agent that attends every customer conversation and becomes your team’s searchable institutional memory.