Copilot Semantic Index Data Refresh

Copper Contributor

In the old Micrsoft Search there is an incremental indexing process to keep the search index up to date and to remove out dated info.  Any idea how does it work with the new Semantic Index (Global Tenant and User Semantic Index) for copilot? 

3 Replies
With Copilot and Semantic Index we move from old word to a newest. Forgot incremental indexing. Semantic Index used different LLM and they are update continuously with new data.

A small video from Microsoft :
Thanks for the reply.
they are update continuously with new data. >> any idea where to find the source of this info?
You can find lot of information on Microsoft Learn.
You should understand that Copilot is not an index. This is how M365 Copilot work :
1) Copilot receives an input prompt from a user in an app, such as Word or PowerPoint.
2) Copilot then pre-processes the input prompt through an approach called grounding, which improves the specificity of the prompt, to help you get answers that are relevant and actionable to your specific task. The prompt can include text from input files or other content discovered by Copilot, and Copilot sends this prompt to the LLM for processing. Copilot only accesses data that an individual user has existing access to, based on, for example, existing Microsoft 365 role-based access controls.
3) Copilot takes the response from the LLM and post-processes it. This post-processing includes other grounding calls to Microsoft Graph, responsible AI checks, security, compliance and privacy reviews, and command generation.
4) Copilot returns the response to the app, where the user can review and assess the response.

If you used a simple index que anwser will be always the same. With Copilot more you will have data more LLM will be efficient. So index information is a little part of Copilot.