Receipt and Invoice AI - Now available in Public Preview!

Help Document Understanding
44

Amazing feature! Eagerly awaiting to use this :grin:

4 Likes
53

Hi @Anupam_Mittal I have included a couple of sample workflows at the bottom of the post, hope it helps!

Alex.

4 Likes
61

Hi @anhtt47 see the original post, I have added sample workflow at the bottom.

Let me know if you run into any issues.
Thanks,
Alex.

5 Likes
65

Hi Team,

Where can I see the saved extracted data?

TIA…

4 Likes
66

Great timing. I was just exploring options for some invoice processing.

4 Likes
67

Hi @alexcabuz,
Can you please tell us, whether we can use these activities in production or not?
And facing some issues while getting invoice number, if the invoice field is in the name of ‘Reg.No’ or ‘Invoice’ in scanned documents.
If the file contains more the 2 document pages it shows some file exceeds error.
can you please suggest some solution to this?

Thanks and regards,
Ramya.

5 Likes
71

Hi @alexcabuz

Awesome that works perfect.
i want to add new fields to be extracted from invoice but after declaring the fields in taxonomy manager and using it in configure extractor i still can extract the newly added fields.
Any suggestions on these.

Can you also help us in knowing some info like
What is the file size limit these activity can process ?

Thanks

4 Likes
73

Hi,

I have the dataset from the export extraction results. But I have processed a multipage PDF file and it only outputs the first-page result when I use write range (excel).

3 Likes
74

Awesome. Does this require GPUs for processing ?

4 Likes
76

hi,

i can’t open this two endpoint link,shows as below:
image
Also, i downloaded your demo script, but no output shows like below:
image
i really want to test this AI function, could you help to answer my questions above?
thanks!

4 Likes
83

Everyone, make sure to check out the example added by @alexcabuz in the first post :slight_smile:

5 Likes
86

@loginerror thank you
Its done, its work, but I want to develop anather service FR langage

4 Likes
88

Hi @Hyungsoo_Kim, it only requires the Extractor engine activity which you can find on the beta feed.

Cheers,
Alex.

5 Likes
99

Thanks! Its a great product. In one of my examples I was testing, I used Bill to Address and Ship to address for invoices. The fields were not detected for billing address and shipping address. I am sure these type of deviations can be there. How to make the tool intelligent enough to behave properly in these deviations

7 Likes
107

Great work UiPath team, this feature will be really useful! In testing the model it was also very accurate :slight_smile:

Looking forward to seeing it’s capability to retrain the model as data is being processed. I will also be looking forward to the ability to create custom fields for specific use cases, such as identification of particular serial numbers or key items within documents.

Keep up the great work!

7 Likes
110

I have just read this blog article comparing the UIPath and other solutions. And I was wondering if anybody had tried extracting US receipts. According to the author, it should work really well…

5 Likes
111

Hi @Esther_Zilber,
Welcome to UiPath Community Forum. Yes I have tried the receipts with English . The microsoft,Microsoft Azure and Abbyy OCRs. It working well. depend upon the quality of scanned receipts.

Regards
Balamurugan.S

5 Likes
112

Hello @balupad14, thank you for the comment and clarification. I am fairly new to the RPA so any guidance is welcome! Have a nice day.

2 Likes
113

You don’t need to access URL. I guess that the URL is for the UiPath activity to connect to the Invoice and Receipt cloud AI engines.

5 Likes
114

got it, thank you

2 Likes