Issue in line item extraction from invoices

i have a invoce.i want to extract data from that invoice.in invoices line items are not fix.line item may very anytime.how can i extract information from that line item.(sometime 1 line item come and some time 3 line item appears)

1 Like

Can you send the two different pdf files???

1 Like

sorry sir…i dont send send this.

Okay can you send two example samples as a string or else screenshot?

ok.
image
somtimes only 1 product id comes and sometimes 3.

Is it PDF Format ?

Run the loop over the text file.

  1. Read line by line
  2. Put the condition if lline.contains(“Product ID”)
  3. Very next line will be your first entry
  4. Append this line into the array.
  5. Run your loop till the line contains product.
1 Like

thank u sir…
i have one more pdf.
image
sometimes only 1 sl no come and sometimes multiple sl no comes.in this case what can i do.??

Please share your text file data in form of the text then I will be able to share the code sir.

Hello @arijit1213

Have you defined the Item field as multiple values in the Taxonomy manager? that should allow you to extract the items and configure the extractors to use UiPath Invoice ML model for those items

Having multiple values in Taxonomy manager doesn’t help you extracting multiple lines automatically. Only if you have user validation you’re allowed to make another line.
I’m also struggling making it dynamically and have no solution, yet.

Yes, of course. But, it doesn’t seem to read all the line items

any solution u found ? i am also wanting to implement a good pretrained model solution for invoices.

Hello, Did anyone find a solution to this?

I am will look be trying to use document understanding for an invoice which contains multiple line items on the invoice?

Thanks in advance

I am also facing same issue in Data Extract for multiple line items with multiple pages

Hi everyone. For an invoice extraction I recommend using the Document Understanding Process found in the templates in UiPath Studio (Version 21.4.4 or higher).
Please use the OOTB Invoice model or retrain it to better suite your needs. Using custom regex is NOT a scalable solution sadly.

If you need more information, I’d gladly help.

Hello!

Do you know how to extract items from an invoice image in pdf format? However, the number of items varies, sometimes there is 1 item, sometimes there are 6 items. (The table is as big as the number of items it contains.)

I would like to solve it with 1 robot to recognize it. The rest of the invoice image is standard.

Thanks in advance!

1 Like

Hi Sandor, it’s quite easy using our in-house ML model. Try downloading the DU Process under the templates in studio and you will be able to solve this.

Hello Sandor,

This is a highly prioritized capability of the machine learning extractor and you can use the out of the box invoice ML package for this, where you can extract different counts of line items.

Here’s how you can deploy the package