In Document Intelligence Studio Layout run analysis is combining two columns in one how can i fix it?

Question

I am using this pdf to test document intelligence studio layout

but when I review the analysis results, I noticed in the result table Price and Per Square columns are combine see screenshot

but in the original pdf are two separate columns, is there a way I can fix this either in Document Intelligence studio or when using the API?

Answer 1

Hello Ali Cruz,

Welcome to the Microsoft Q&A and thank you for posting your questions here.

I understand that you observed in Document Intelligence Studio Layout run analysis is combining two columns to one and you need how you can fix it.

If the PDF doesn’t have clear separators (lines, spacing, or borders), the model sometimes merges two adjacent columns into one cell exactly like you’re seeing with Price and Per Square.

You can’t manually “fix” the Layout model’s column detection in Studio. The options are:

• Try document model (instead of just layout).
• Improve table formatting in the PDF by adding clear separators.
• Split values programmatically after extraction.
• Train a custom model if you need consistent separation:

In Document Intelligence Studio, you can build a custom extraction model (using “custom extraction” instead of just layout). You can label Price and Per Square separately in training data. The trained model will then learn to separate them correctly, even if the Layout model merges them.

I hope this is helpful! Do not hesitate to let me know if you have any other questions or clarifications.

---

Please don't forget to close up the thread here by upvoting and accept it as an answer if it is helpful.