Problem on not correctly detecting VendorName

Hi @Dhruvin Vachhani,

Thank you for reaching out to Microsoft Q&A forum!

To extract data from PDFs using Document Intelligence, you can use the pre-built models or create your own custom extraction model. If the pre-built models are not accurately detecting the vendor name, you can try creating a custom model that is trained on your specific vendor documents.

To create a custom model, you can train the model using your own vendor documents and labels to improve the accuracy of the model. Here are the general steps to create a custom model:

1. Collect a set of vendor documents that represent the type of documents you want to extract data from.
2. Label the data in the documents to identify the fields you want to extract, such as the vendor name.
3. Train the model using the labelled data.
4. Test the model to ensure it is accurately extracting the data.
5. Use the model to extract data from new vendor documents.

For more info:

• Custom model overview.
• Build and train a custom extraction model.

If you still encounter any errors, please share your sample documents with us on the private message and we will work on resolving the issue.

I hope this helps. Thank you.

Do not forget to click Accept Answer and Yes for was this answer helpful.