Connect Yogi MCP with M365 Copilot

This guide will walk you through how to connect your Yogi account to M365 Copilot using the Yogi MCP, which enables you to ask questions about your customer feedback data in natural language, within the AI tools you already work in.

• Model Context Protocol is a standard that lets AI tools connect to external data sources.

• With the Yogi MCP, your AI tool can access your Yogi data and interact with your customer feedback data.

Install the Yogi MCP on Copilot

Yogi connects to Microsoft 365 Copilot through a Copilot Studio agent — there's no direct connector yet. One person with a Copilot Studio license builds the agent and adds Yogi as a tool, then an admin approves it. After that, Yogi appears in the Agent Store for everyone, and each person signs in with their own Yogi account.

Agent builder steps (one-time setup)

1. In Copilot Studio, go to Create → New agent.

2. From the agent's Tools page, click Add a tool → New tool → Model Context Protocol.

   

3. Fill in the server details:

   1. Server name: Yogi

   2. Server description: Consumer feedback and voice-of-customer analysis across reviews, ratings, and social. Use for questions about what customers say about products, brands, sentiment, and themes.

   3. Server URL: https://api.meetyogi.com/mcp

4. Under Authentication, select OAuth 2.0, then choose Dynamic discovery. Copilot Studio will find the right endpoints and register itself automatically — you don't need a client ID or secret. Click Create.

   

5. In the Add tool dialog, under "Not connected," choose Create new connection. A sign-in window opens — log in to Yogi with your app.meetyogi.com credentials and authorize. The connection label should flip to your account.

   

6. Confirm the tools came through. Copilot Studio pulls in Yogi's tools automatically, and they should be listed on the Yogi tool's settings page with their names and descriptions. This is the checkpoint that matters — if the tools are there, the connection is working.

   

   

7. Set Credentials to use → End user credentials. This means each person signs in with their own Yogi account and sees only the data they have access to. The alternative, maker-provided credentials, runs Yogi as your account for everyone in the org.

8. Test in the test pane. Ask something Yogi should answer, like "Show me the top themes for [product] in the last month."

9. Publish the agent, then go to Channels and select Microsoft 365 Copilot. Use Edit details to set how the agent appears, then save to submit it for admin review.

10. An admin approves it in the Microsoft 365 admin center under Agents → All agents → Requests. Once approved, the agent appears in the Agent Store under "Built by your org," where admins assign access or deploy it to specific people and groups.

Individual user steps:

1. In Microsoft 365 Copilot, add your org's Yogi agent from the left pane. You'll find it in the Agent Store under "Built by your org." If it isn't there, ask your admin to grant you access.

2. Ask the agent a question about your customer feedback. The first time, you'll be prompted to connect — click Open Connection Manager.

3. In the Connection Manager window, click Connect, then the three-dot menu, then Add new connection.

4. Sign in with your app.meetyogi.com credentials and authorize.

5. Return to your chat and retry your question. From then on, you can ask Yogi questions in any chat with the agent.