Forum Discussion
Searching for a simple guide to index SharePoint and publish an agent in Foundry
Hey all,
Does anyone have a good guide or best practices for this setup in Foundry?
- SharePoint as data source
- GPT model (document + image indexing, ideally vectorized/embeddings)
- Create an Agent an Share the Agent
- Restrict access to Agent to specific users/groups only
Looking for tutorials, examples, or real-world setups.
Thanks!
1 Reply
The simplest supported setup is to start with Foundry’s SharePoint tool, rather than build your own index. Add SharePoint under the agent’s knowledge tools, create a connection to the site or folder, and test with users who already have access. It uses delegated identity, so SharePoint permissions remain the security boundary and results are trimmed per user.
One caveat: the SharePoint tool currently retrieves text extracts only; it does not understand images or charts. If image OCR or your own embeddings are required, build an Azure AI Search index from SharePoint, add OCR and vectorization in the indexer skillset, and connect that index to the agent instead.
When ready, publish the agent. Grant the intended Entra security group the invoke role on that specific published agent resource, then grant the published agent identity access to any downstream Azure resources it uses. Test with one allowed and one denied account before rollout.
