In this post, I will try to explain the inner workings of Copilot and how we can preprocess prompts with additional data to generate better responses from the underlying Large Language Models. I will finish it off with an example of how this theory could be manifested and used in practice. 1. Background 1.1 Dynamics Customer Service Copilot Surly you have heard about Dynamics Customer Service Copilot by now. The product comes in the way of a chat side pane integrated within Dynamics and lets you generate email replies, summarize cases, and generate responses based on knowledge articles.

Recently I had the opportunity to click around in a Legacy System. This got me thinking about the evolution of IT systems and how legacy systems still are in use even though there is newer technology out there. Changing IT systems truly is a painful thing, and it is even worse for a business-critical system. This article aims to first compare Dataverse to legacy systems and then advise on how you can improve your legacy system by making some architectural changes.

Background In my last blog post How to implement extended audit logging in dataverse Part 1 I showed how we can use the out of box Microsoft Purview/Office logging to send Retreieve events in Dataverse and then use Power BI to visualize this information back to users in Dataverse. However, the entire process of setting everything up was quite cumbersome and involved a lot of moving parts. So I was in search for something simpler.