Exploratory Testing with Teams Facilitator

This week I noticed I had access to Facilitator in Microsoft Teams, an AI agent that takes real-time notes. When it's enabled, you can ask it questions both during and after the meeting, and it answers based on what's been said in the call or posted in the meeting chat.

I parachuted into a project yesterday to do a round of exploratory testing in production on a system that had just gone live for the first time, to verify the integration points. I was going to use Teams for screen recording anyway, and then I had the idea of using Facilitator to write my notes.

The rough steps were:

Very well! Not having to take notes manually saved a lot of time and left more time for exploring.

My template uses PQIP (Problems, Questions, Ideas, Praise). I briefly described this at the start, and anything I wanted to flag to the team I said out loud, e.g. "Facilitator, make sure you note this as a problem." It also did a great job of picking out points for each section when I didn't spell them out.

The one thing I lost was a record of where I was up to. I got pulled into a couple of side quests in other Teams chats. Normally I'd check the last point in my notes and carry on, but this time I think I covered some things twice. In hindsight, I should have asked Facilitator where I was up to!

I compared the notes side by side with the transcript, and it captured everything important. There were a couple of issues, though: it left out all the URLs I'd shared, and the markdown formatting wasn't right. I fed the notes to Claude afterwards to sort them out.

Overall, it was a big time saver. I'll definitely be doing this again.