Hello, I’m on enterprise plan with consumables. I created a project in DU that has 20 document types (all fields configured using uploaded batches), I’m trying to figure out how classifiers defined here work with the classification scope in studio. how to configure the classification scope so that it uses the classifiers in DU rather than the classification json.
or the classification scope is no longer needed? then how should I configure the extraction scope in terms of classifiers?
@Anil_G would you please let me know how to use the semi-structured form AI extractors? I’ve been only using the form AI by creating the end points, but the semi-structured form AI don’t have a publish button…
I haven’t done anything with pipeline and found the documentation confusing, is there a video or link that demonstrate this piece? thank you!
and here is the answer I got from chatgpt4:
In UiPath Document Understanding for handling semi-structured documents, there isn’t a traditional “publish” button as you might see in other parts of the UiPath platform. Instead, the process involves a series of steps to set up and convert your document types for use with the AI models available in UiPath’s AI Center.
Here’s a streamlined approach to getting your endpoint in the context of UiPath Document Understanding:
Creation and Conversion: Start by creating a document type using Forms AI within your Document Understanding project. If you begin with a Forms AI project and need to handle semi-structured documents, you’ll need to convert this document type to a Semi-Structured Document Type. This conversion allows the use of deeper machine learning models and more complex extraction capabilities tailored for semi-structured documents.
Using AI Center: Once you have your document type set up and configured, integrate it with UiPath’s AI Center. In AI Center, you can deploy machine learning models as skills. These ML skills can then be used within your Document Understanding processes.
Endpoint Creation: To create an endpoint, you first need to set up an ML Skill in the AI Center. After setting up the skill, you publish it in AI Center, which provides you with an endpoint. This endpoint can be used in your Document Understanding workflow to apply the ML model to your documents.
Integration in Workflows: Finally, integrate this endpoint within your UiPath Studio by using activities like ML Skill Activity or Document Understanding activities that connect to your ML models in AI Center.
This process leverages the ML capabilities of AI Center to process documents, and there’s no direct “publish” button in the traditional sense within the Document Understanding framework. Instead, endpoints are managed through AI Center where skills are published and maintained.
For detailed guidance on setting up and using these features in UiPath, the Document Understanding and AI Center documentation provides comprehensive resources. You might want to explore more about how to manage document types and integrate ML models directly from the official UiPath documentation.
