document_ai_warehouse_processing_python

Using Document AI Warehouse Python Client Library to run common operations on Document AI Warehouse

Overview

This project demonstrates how to perform common actions on Document AI Warehouse through the API. dw_processing.ipynb uses DocumentWarehouseUtils.py for abstraction and readability. It is recommended to look at the code provided in the utils python files.

1. Create document & folder Schema.
2. Create a folder using schema created in step #1.
3. Create a document using schema created in step #1 using inline raw document & set property values.
4. Create a document using schema created in step #1 using document stored in gcs & embed DocumentAI processor output alongwith.
5. Link document created in step #4 to the folder
6. Search document
7. Clean-up

Prerequisites

1. Please ensure that you have a Document AI Warehouse instance in your project. You can follow this quickstart to complete the setup.
2. Create a Document AI Invoice processor and update the DOCAI_PROCESSOR_ID variable below.
3. If you are using a Vertex AI Workbench Managed Notebook, ensure to grant the following roles:

• roles/contentwarehouse.documentAdmin
• roles/documentai.apiUser

If you are using your own dev environment please ensure to grant the specified permissions to the identity.

Setup

1. Install dependencies mentioned in requirements.txt

pip install -r requirements.txt

2. Open dw_processing.ipynb and follow the step-by-step guide.