How do i add new documents to my Microsoft Graph rag solution accelerator deployed on Azure?

Question

Hi, I am working on increasing the amount of files in the Microsoft Graph Rag solution accelerator index but there is no suggestion on how to do it in this repository https://github.com/Azure-Samples/graphrag-accelerator and it appears to be archived. However, this repository https://github.com/microsoft/graphrag shows how an incremental indexing can be done but this does not apply to the solution accelerator I believe. Could you please suggest how this can be achieved ?

Answer 1

Hi Oluwasogo, Kehinde (Data Evidence and Performance) ,

Thank you for reaching out on Microsoft Q&A.

The GraphRAG Solution Accelerator you deployed is archived and does not support incremental indexing. To add new documents, you can:

1. Upload new files to the data API:

      curl -X POST "$APIM_BASE/data" \
   

-H "Ocp-Apim-Subscription-Key: $KEY"
-F "container_name=<your-input-container>"
-F "overwrite=true"
-F "files=@/path/to/new1.pdf"

   
   **Re-index** the same index:
   
   ```yaml
   curl -X POST "$APIM_BASE/index" \
  -H "Ocp-Apim-Subscription-Key: $KEY" \
  -H "Content-Type: application/json" \
  -d '{"storage_container_name":"<your-input-container>","index_container_name":"<your-index-name>"}'

Check status:

   curl "$APIM_BASE/index/status/<your-index-name>" \
  -H "Ocp-Apim-Subscription-Key: $KEY"

For incremental (delta-only) updates, use the maintained microsoft/graphrag library and run:

graphrag update --root /path/to/workspace

References:

GraphRAG Accelerator (archived)

GraphRAG Library

CLI Documentation

DeepWiki User Guide

Microsoft Research Blog

I hope this helps. If the information is useful, please accept the answer so it can assist other community members.