Doc2Edi

First Engine:

Extracting Data from Document PDF Files Using GCP and Google Document AI

Steps for Engine-1

1- create new project on GCloud: https://codelabs.developers.google.com/codelabs/docai-ocr-python#1 2- create service account for this project: https://console.cloud.google.com/iam-admin/iam?walkthrough_id=iam--create-service-account&project=oko2-386015

NOTE: these roles should be selected for this service account: Document AI Administrator Document AI API User

3- create a processor for my project: https://cloud.google.com/document-ai/docs/create-processor?_ga=2.114028359.-1141760794.1683471749 NOTE: for this project I used Form Parser from G Doc AI

4- cloned this repo: https://github.com/anirbankonar123/documentai 5- added my auth key location into my windows env variables as: GOOGLE_APPLICATION_CREDENTIALS=’path to json file’ 6- copied my project and processor info into my local code 7- for each PDf file, the code ran and created one json file and multiple CSV files.

python doc_ai_table.py --pdf <pdf_path> --folder <output_path>

Second Engine:

Converting Data into EDI 211 Transmission Files

Steps for Engine-2

1- Edi Fields Finder 2- Find Best Edi Field & Process 3- DQ Checks & Valid Values 4- Put in EDI 211 Format

NOTE: required fields for EDI 211: • Shipment ID number • Date and time (of pick-up / delivery) • Status report request (upon delivery) • Business instructions • Handling requirements • Bill of lading rates and charges • Lading quantity / weight / freight Class / value • Contact information

Resources:

1. Document AI tutorial: https://cloud.google.com/document-ai/
2. How to setup my Doc AI project: https://codelabs.developers.google.com/codelabs/docai-ocr-python#0
3. repo used for the first engine is from: https://github.com/anirbankonar123/documentai