Choose a starting point
Start here
You don’t need to read everything. Choose the problem in front of you and follow a short path through the work.
Make sense of Microsoft AI
Understand the result. Then the trade-offs.
Start with the benchmark.
See how five models responded to the same prompt, with the method and limits kept in view.
1. Read the Copilot Cowork benchmark
Then look at the operating side: how an admin can review the first month of Copilot Cowork consumption.
Build better automations
Get the details working together.
Start with a dependable foundation.
Learn the role of environment variables when you move a Power Automate build between environments. This guide was published in 2020; keep that context in mind as product screens evolve.
1. Explore deployment parameters
For a smaller practical task, follow the updated date-difference guide.
Connect AI to your data
Watch the setup. Keep the reference.
Bring Cowork and Dataverse together.
Start with the written guide for prerequisites and setup context, follow the video, then return to the article for files and troubleshooting.
Looking for something specific?
The article library keeps the wider archive in one place, including older Power Automate tutorials and newer Microsoft AI field work.
Find the guide for your next step.
Use these topic paths when you know what you want to build or troubleshoot.
Power Automate
Start with dateDifference syntax, then extracting days, hours, minutes and seconds. Read the examples’ limits before reusing the expressions.
For Excel work, explore removing duplicate values with union() and using a dynamic Excel file.
Copilot Studio
Start with Microsoft Learn MCP setup, continue to publishing the docs agent in Teams, then explore customer-operations hooks.
Follow each guide’s prerequisites and current-guidance notes; these are separate tasks with different setup requirements.
Cowork operations
Begin with Cowork billing setup, review consumption and adoption, then connect the Dataverse plugin.
The consumption walkthrough includes a specific month’s results; use that context when comparing your own environment.