SAP Community - Document Information Extraction

Document Information Extraction - Template Inaccurate

2023-10-11T17:18:17+02:00

Hi All,

I want to use the templates feature of Document Information Extraction service for an custom invoice which has purchase order header and item details. The invoice looks like below screenshot.

The issue I am facing with a custom template is that, the line items are being read inaccurately. Even though I have annotated the documents properly for the position of line items and fields as below.

I have uploaded up to 4 documents with same annotations. Still the Extraction results comes like how its shown below.

The line items are ready fully inaccurately. Could anyone please help in addressing the issue here?

Thanks & Regards,

Vishal Raghav

SAP Document Information Extraction for Payment Order using excel

2023-11-14T15:27:21+01:00

Hello Expert,

We have requirement for reading the content of the excel file using dox. At present its only support for document type of payment order with excel file upload. In our case its working fine but we have added some custom fields and could not able to train the model in UI as no preview is available. Please help us to know how can we train model in UI for excel file.

Document Information Extraction : Purpose of Documents and Templates

2023-11-20T07:53:36+01:00

I have just installed the Document Information Extraction. Wondering what the purpose is for Documents and Templates. If I have 3 different Supplier Invoices do I create 3 documents and then tie all of those into 1 Template ?

I have at the moment uploaded 3 documents - when uploading a document you must choose a schema and a template. You can see from the screenshots that my documents have a template "PetersTest". But when I look at the Template only 1 document is there and this is because I specially added it on that Template tile. I was expecting it to have all 3 documents.

SAP CIM and DOX - how are they known to each other?.

2023-11-22T01:57:27+01:00

I've learnt from Jie Xiong that SAP CIM is a DOX powered application. I can see that now in the Architecture diagram

I had installed SAP CIM and for a few days was testing the OCR scanning of Supplier Invoices. But I was not happy with the OCR as it was missing elements like the Price and UOM - unless I had a simple Invoice layout it was happy with. Then I stumbled upon Document Information Extraction (DOX).

So I had SAP CIM working in my BTP without even having DOX installed in my BTP !!. So in fact I did not even know that you had to install DOX - well in fact it is not required as it works (poorly) without it.

So I then installed DOX.

During CIM install there was no mention of DOX or Document Information Extraction in the guide (4N6_Set-Up_EN_XX.pdf - ok it is mentioned briefly in the Help). In the set up of DOX - I used the BTP Booster there is no mention of CIM nor in its help

So how do the 2 know about each other ?. If I had SAP CIM DEV, QAS & PROD and SAP DOX DEV & PROD how would they know which one to check ?

Audience Validation Failed Persists Document Information Extraction

2023-11-22T10:50:12+01:00

I am trying to make an API call with the token generated with the link: https://b3cxxxxxxxx.authentication.us10.hana.ondemand.com/oauth/token?grant_type=client_credentials and my ClientID and ClientSecret as per my Document Information Extraction Secret Key generated on BTP.

The token that I obtained was checked by me on the subject of the audience. And the data in the payload of the token was correct and corresponded to my service key json data.

However, when I try the token in https://aiservices-dox.cfapps.us10.hana.ondemand.com/document-information-extraction/v1/, I get the 401 Error Audience validation failed. I tried refreshing the token, refreshing the browser, deleting cookies, and NOTHING WORKED. One day, the token was accepted and the API call was successful; the next day, the new token failed completely randomly and without regard for any rationale. Please advise on how to resolve the bug. Thanks!

P.S Every time I insert my token with Bearer before it. The Audience Failed error is appearing and disappearing regardless. Yesterday I had 40 attemptions to execute an API and it didnt worked. On the 41st attemption it worked. Today I have the problem again with a newly generated token. Apparently, there is a bug. Please advise on how to fix it. Thanks!

Processing custom documents through SAP-DIE Tool

2023-11-23T11:55:59+01:00

Hi,

1. Is there a way I can process the Settlement Statement, Imbalance summary report & Monthly activity reports documents through SAP-DIE Tool other than the default document types already available in the tool which are (Invoice, Payment Advice & Purchase Order)?

2. How, where/when can the custom document type can be selected? (Purpose of the Custom document type).

Document information extraction

2023-12-07T11:24:44+01:00

Hi, I would like to know if there is historical data saved somewhere in DOX. Or is DOX model trained based on historical data?

DOX further procedure

2023-12-07T13:04:36+01:00

Hi, what happens after the information is extracted from a document (eg. invoice)? What happens after downloading the extraction in either csv, json or txt? Can you please explain the next step?

Thank you,

Medha

Document Information Extraction Consumption Quotes

2023-12-07T13:51:50+01:00

Hello! According to the sap provided information, for the trial BTP subbacount, and the trial subscription to Document Information Extraction, I am only allowed to upload 50 documents per 30 days. How comes, I managed to process 749 documents in the course of 2 days? image.png Please find the screenshots from my Usage Analytics in the attachment. I would appreciate some help to understand my consumption limits. Thank you!

Can I train my own Document Information Extraction model?

2023-12-13T16:22:14+01:00

I am currently trying to figure out a way to train my own Document Information Extraction model. I work with pharmaceutical batch records and need to extract the information. Creating a template for them in the standard Document Information Extraction app wouldn't work because they are all quite diverse. Can someone recommend to me the best approach to take?

Within SAP, where can I develop this application that will allow it to access foundational models? Can I train, test, and verify the model within said development environment? Is there a way to leverage generative AI to aid me?

In a OpenSAP course I did, the lecturers mentioned that we'd be able to further train existing Foundational Models to suite our use cases. Is that available yet, and if so, where is it situated?

How do I access the Generative AI hub? Does it form part of the AI Launchpad?

I am new to the SAP space and find it challenging to navigate all the different products and services.

BTP Document Information Extraction Service - Retrieval of deleted document data

2024-01-02T08:26:00+01:00

Hi,

We are using SAP BTP - Document Information extraction service in which we have created templates & uploaded scanned documents against the templates as well.

Everything was working fine & data was being scanned as well.

Subsequently we observed documents uploaded earlier are automatically being deleted from SAP BTP.

For .e.g, document uploaded on 01st Sept 2023 is no longer visible in BTP.

https://help.sap.com/docs/document-information-extraction/document-information-extraction/data-protection-and-privacy

We wanted to understand if there is a way to restore the documents in SAP BTP Document extraction service

Regards,

Gavin.

SAP Document extraction service not working for uploads

2024-02-03T03:55:37.499000+01:00

HI Team,
I have tried exploring the Document extraction services(premium) from postman to build some apis to upload the document to extract some information. All get calls related to document extraction services are working in postman and code. But upload functionality not working in postman and code. In the postman input, in form am passing file and client id which are required params. But am getting following error

{

"error": {

"code": "E95",

"message": "Required form-data not provided.",

"details": [

{

"code": "0",

"message": "Missing parameter: options"

}

]

}

Any help is greatly appreciated.

Can I use DOX Custom Schemas (w/ AI functionality) with the DOX SDK in SAP Build Process Automation

2024-03-15T10:29:08.739000+01:00

Hi,

Can I use DOX Custom Schemas (w/ AI functionality) with the DOX SDK in SAP Build Process Automation?

I know that premium version is needed.

If yes, how can I enhance the DOX SDK.

Thanks !

Tillbtp

Problem with data types of DOX Payment Advice Data Type in SAP Build Process Automation

2024-03-19T16:17:39.924000+01:00

Hi,

I am trying to process a payment advice automatically by using SAP BPA and Document Information Extraction with the Extract Data (Pre-trained Model) for Payment Advices.

I always get an error log mentioning a discrepancy in data types. From what I understand, DOX's is giving the item 'netAmount' as a number as output but the netAmount.value in DOX SDK is actually formatted as a string.

Attached some screenshot for better understanding.

Output of automation job:

Error Log:

I think the discrepancy is already coming from the data types before.

Best wishes and kind regards

Till

SAP BTP Process Automation how to Extract Invoice Details

2024-04-02T12:55:21.440000+02:00

Hello SAP Community,

I'm trying to build my first process in SAP BTP Build Process Automation following the Tutorial

Invoice Processing and Approval Using SAP Build Process Automation

The problem is in my developer environment I'm missing the option to add the "Document Template" (see the Screencopies - first one is from the example and the second one is from build studio)

I also checked the dependencies that the Document Information Extraction SDK was added to my project

Would be great if someone could get me back on track.

Thank you

Andreas

AI Document information extraction

2024-04-09T14:53:06.863000+02:00

Hi there,

I am currently using the AI Document information extraction Service and was wondering which LLM is used for this service. I couldn't find any information on this and it would be great if someone would help me. Thanks!

Best regards,

Julian

Document Information Extraction Premium using the UI only - Process the Extraction Result (JSON)

2024-04-10T14:48:56.044000+02:00

Hi all,

We are looking into the functionality of Document Information Extraction (premium) on the SAP Business Technology Platform . We got the UI version working on a BTP trial environment and all looks promising, but not (yet) perfect. Since we receive lots of documents from lots of different parties, our intention at first would be to use only the UI to upload the documents. Uploading the documents and extraction the data works pretty good, but we now are trying to figure out how we can process the Extraction Result in a user friendly manner. So far, the only available option we've found is to download the Extraction Result in the requested format (JSON). The JSON file is then downloaded to my windows downloads folder.

Are there any other options to process the Extraction Result from the UI so it is pushed to / loaded in our SAP backend in a more automated manner. At this time, we are not (yet) interested to process the documents using the available APIs.

Document type - Document Information extraction.

2024-04-11T17:39:17.438000+02:00

I understand document Information extraction would create a document of document type custom, is it possible to push this document to a backend S4HANA system ?

SAP Build process Automation - Document Extraction Service: Missing Rights?

2024-04-12T09:33:08.931000+02:00

Hi there,

I am currently trying to Automate a process using the Build Process Automation Service. I was planning to use the Document Information Extraction Service in this Automation, but for some reason it doesn't show up if I am looking for it.

Normally I should be able to select "Document Template" but if I use the Build Process Automation Service, its just not there.

Same thing here: I should be able to use an "Extract Data" Activity, but it's not there:

I assume its because of missing rights, but I don't know which rights could be missing. I am able to use the Document Extraction Service in the BTP, but not inside SAP Build

Does anyone know what might be the reason for all of this? Thank you!

Julian

Edit: If I try to run an imported project with an "Extract data"-Action I get this Error:

{

"uid": "XXXX",

"name": "Extract Invoice Data",

"type": "automation",

"instanceUid": "XXXX",

"packageUid": "XXXX",

"packageVersionUid": "XXXX",

"attended": true,

"duration": 1,

"status": "Failed",

"code": "KO",

"label": "[requestDoxToken] Error on content received: {\"name\":\"requestDoxToken\",\"tid\":86,\"code\":\"InvalidCommand\",\"label\":\"No bean named 'requestDoxToken' available\"}",

"exception": "irpa_core.error.Error",

"stack": "Error: [requestDoxToken] Error on content received: {\"name\":\"requestDoxToken\",\"tid\":86,\"code\":\"InvalidCommand\",\"label\":\"No bean named 'requestDoxToken' available\"}\n at C:\\Users\\I7XXXX at o.receiveMessage (C:\\Users\\I7XXXX at c (C:\\Users\\I7XXXX at u (C:\\Users\\IXXX at l (C:\\Users\\I7XXXX at C:\\Users\\I74XXXX at o.p [as onEvent] (C:\\Users\\I7XXXX at process.<anonymous> (C:\\Users\\I7XXXX at process.emit (node:events:513:28)\n at emit (node:internal/child_process:937:14)"

}

XXXX means real content substituted with "XXXX"

How to export DOX templates from SAP RPA to SAP BUILD

2024-05-17T12:39:38.697000+02:00

Hi everyone,

I am moving to SAP BUILD after 2,5 years of using SAP RPA and as I perfectly managed to import my projects, I am facing some difficulties with those using DOX templates.

As far as I understood, SAP RPA and SAP BUILD cannot coexist in the same subaccount in BTP cockpit, so I installed SAP BUILD in another subaccount. What should I do to have my DOX templates available in my imported projects? Should I set up DOX in a certain way in the new subaccount, using the same service key maybe? Export/import them manually? Please note that I cannot export any of my templates on DOX UI, each attempt resulting in a pop up saying:

Thanks beforehand for your help.

BR,

LM.