i wanted to seek your guidance on a particular case that my colleague Gagan and I are working on regarding Copilot Studio.
We have a customer who wants to create agents and utilize them as a managed service for their partners. Here is a brief description of the use case: The customer is using Copilot Studio to develop a Copilot agent app and deploy it in Teams. This app serves as an additional option for their clients who use our ITSM platform. Being the service providers, they want the capability to develop the app within Customer's tenant and then publish/deploy it on their clients’ Teams.
We have a few questions regarding the infrastructure requirements for this setup. Specifically, we need to understand the tenant-to-tenant connectivity, Teams setup on the clients’ side, and other relevant details. Additionally, we need to know how the cost will be calculated. Since the customer follows a service provider model, they need to consume all the costs while also informing the client about their consumption-related costs, possibly on a monthly basis.
Moreover, we are curious if there is a need for separate environments for each client solely for tracking costs.
Can you clarify if details (knowledge) from 'client' needs to be exposed to agents from 'customer'? If so, this will be a security risk consideration for the 'client'.
It will be great to understand the exact scenario. Why can't the agent be developed by customer and deployed in an environment of the 'client'?
If customer will host it and pay the cost, then Azure AI Foundry might be a cheaper alternative compared to consuming Copilot Studio credits.
Was this reply helpful?YesNo
Under review
Thank you for your reply! To ensure a great experience for everyone, your content is awaiting approval by our Community Managers. Please check back later.
Share
Report
All responses (
Answers (
646
