Problems setting up Document AI

Google Cloud Build with AI AI Solutions
1

Hi all.
I’m trying to set up Document AI in my Google cloud console and make API calls to use the Document OCR processor.

I’m getting somewhat confused with the naming conventions, and want to double check I’ve set things up properly at the google cloud end, before troubleshooting at my development end.

  1. I’ve enabled Cloud Document AI API

  2. I’ve setup service credentials as Document AI API User

  3. I’ve downloaded my json key and added to my local production environment

  4. Now, this is where it gets confusing, To enable a processor, you have to navigate to a page, which ISNT linked from the Cloud Document AI API entitled Document AI Overview. This page contains links to enable processors. Finding this page was tricky and strangely only linked from doing a search in the main toolbar. It made me think…are these services one and the same thing? I’m assuming they’re just named differently.

  5. Anyway, I enabled the Document OCR processor and took note of the ID and the region for my production environment.

In my prod environment I have
GOOGLE_APPLICATION_CREDENTIALS (my json service key)
GOOGLE_PROCESSOR_ID (from the enabled Document OCR processor)
GOOGLE_PROJECT_ID
and GOOGLE_LOCATION
I’ve also installed the document AI API client library

Is this correct? Have I missed anything on the google cloud side?

2

Hi @oranges-901
Everything you listed looks correct on the Google Cloud side.
And yes, the processor setup and the API are part of the same service — just accessed through different parts of the console, which can be confusing. At this point, you should be ready to make API calls. If something doesn’t work, feel free to share the specific error.