When we adopt Microsoft 365 Copilot we want all our data to be available for Copilot to use when gathering information and creating content back to us. So we should alwasy have a strategy to use all Microsoft 365 services to the max and move or migrate our data into the Microsoft 365 services.
When I say ALL our data available there are of course limitations for each user on what the user has access to on files and items.
Copilot accesses content and context through Microsoft Graph based on what the user that initiate the request has access to, and the response will therefor only include data created based on what the user has access to.
As we say that Microsoft Graph is what Copilot uses to collect data for modeling, we can also connect other sources from outside our Microsoft tenant into the tenant’s Graph so that we expand the source of data to much more than only Microsoft 365.
Microsoft has many source connectors like Atlassians Confluence both On-prem and Cloud or Fileshare connectors, see a complete list here at learn.microsoft.com.
Connectors have different limitations, but never the less – we are able to actually create a connection between a third-party system or fileshare to Microsoft Graph meaning that we can utualize Microsoft Search for searching content across systems living outside of our Microsoft 365 tenant.
And of course with this capability we also expand what Microsoft Copilot can get hold to of source data to create the perfect result of our prompt that we ask it for.