How do you improve evaluation when working with Microsoft Copilot?

[lucy_3gju]

I have been exploring Microsoft Copilot and wanted to ask people who have tried it in practice.

I am particularly interested in evaluation and real-world use. I would like to compare approaches that people have actually used rather than just theoretical suggestions.

What trade-offs have you found between quality, speed, cost, and reliability?