I’m working on a finance AI project in Copilot Studio where the goal is to build an agent that performs financial analysis and automatically generates variance commentary.
The agent needs to analyze Actuals vs Forecast / Plan / Prior, identify key drivers, and produce commentary to go on our powerpoint slides suitable for finance reporting.
The data that I'm using to provide as knowledge to the agent is quite complex. Its approx. 30mb excel file that basically contains raw data about our sales. I tried to upload the file directly into copilot studio but I'm not getting any accurate responses when I ask it questions. For example how much of X product was sold in Period X, and the idea is have it answer more complicated questions like why did this product do bad/good as compared to last year. All of that information exists in the raw data and our analysts manually dig through it to generate commentaries but I was wondering if this can be made easier with AI involvement?
Any suggestions or advice would be greatly appreciated!
Thanks in advance

Report
All responses (
Answers (