Answered

how to extract data or text from the pdf in the power automate flow

(1) Share

Report

Posted on by CU20121706-0

I am trying to extract Invoice Number, Date, Entity, Vendor Name and Amount/Total from the pdf to store in one of the SharePoint list. I have around 200 types of different pdfs to extract the information from and did lots of research but could not find the right tools or Premium app or any API to use in my power automate flow. I have option for AI builder but as I have more then 200 types of pdf and not possible to build model for each pdfs. Is there any solution that I can use to extract the invoices info from the pdf? I dont mind to purchase any Premium App or API to do this job as long as it fulfill all my needs.

There are some features in different type of PDF as below,
some pdf has image where the invoice number ,dates and amounts is

some pdf has table contain and has invoice info.

each pdf has different format from every vendor.

some has excel file as an invoice and some has jpeg image to extract invoice information.

If anyone please help me to find a right tool or app to install in power automate, would be really appreciated.

Cheers,

Categories:

Building flows

I have the same question (0)

All responses (3)

Answers (1)

Verified answer

Michael E. Gernaey 53,442 Super User 2025 Season 2 on at

Like (1)

Report

Hi

So, its funny as I had a customer who asked essentially the same thing. The issue is, you need to build a model, and you need to feed in all of the samples so that it can. Problem is you aren't going to get what you want based on what you are asking for. At least not easily. Its not specifically Power Automates fault, its really a matter, what it will take to insure that you have the properly trained back end model.

So the answer is... its not simply installing something, so in that sense, the answer is No, you won't just install something and it works. You have to do what Is said above and its a lot of work. Not impossible but not easy.

Was this reply helpful? Yes No
CU20121706-0 2 on at

Like (0)

Report

Hi @FLMike

I got your point that I have to build a model in AI builder for each pdf template. I would like to ask you instead of AI builder is there any other way I can extract invoice data or not in power automate?

Was this reply helpful? Yes No
Suggested answer

takolota1 4,974 Moderator on at

Like (0)

Report

You can extract data to a json object or json array from any type of document with any style with OCR & an MS GPT prompt. See this template: https://community.powerplatform.com/galleries/gallery-posts/?postid=31e67eea-3f73-47b4-95b7-fe4a7b646389

Was this reply helpful? Yes No