Document Intelligence Studio - Multiple people working simultaneously

Benjamin Pinter 0

We have developed an application that enables our staff to train custom extraction models, which are subsequently integrated into our downstream processes. Currently, we are encountering challenges in facilitating parallel training sessions for multiple users. Specifically, each user is unable to maintain an individual set of labels; instead, all users share a common label configuration.

Our objective is to allow multiple team members to train models concurrently, with each user operating with their own distinct label set. This setup would enable the creation of a composite model that integrates the efforts of all users.

Could you provide guidance on configuring Document Intelligence Studio to support this parallel, user-specific labeling and training workflow?

kothapally Snigdha 1,620 Reputation points Microsoft Vendor

2025-02-27T00:02:09.0766667+00:00
Hi Benjamin Pinter

To allow multiple users to concurrently train custom models with their unique label sets in Azure AI Document Intelligence, you can follow this method

Establish individual labeling projects for each user, enabling them to independently define and manage their labels.

Each user can train their model using their labeled data, ensuring isolated label sets and parallel training.

After training, integrate the outputs from all user models in your downstream process. Utilize asynchronous API calls to run models in parallel and merge their results using business rules.

Ensure users have access only to their respective data and labeling projects through Azure’s role-based access control (RBAC).

I hope these helps Thank you!

Share via

Document Intelligence Studio - Multiple people working simultaneously

Your answer