devoteamgcloud / doc-ai-demo

About The Project

This demo illustrates the capabilities of Document AI and Gemini.

Deploy the app locally

• Create a document AI processor , in this case , it is an OCR type processor.

• Create a .env file with the following environment variables:

  PROJECT_ID='YOUR_PROJECT_ID'
  LOCATION='eu'
  PROCESSOR_ID ='YOUR_DOCUMENT_AI_PROCESSOR_ID'

• Create a Dockerfile file

• Build the image locally, cd into your project folder and run.

  docker buildx build .

• Run the docker image using the application default credentials used for authentication.

  docker run -v "Path to you loclal application default credentials i.e application_default_credentials.json":/gcp/creds.json:ro  --env GOOGLE_APPLICATION_CREDENTIALS=/gcp/creds.json -p 8080:8080  "YOUR_DOCKER_IMAGE"

• Access the app via the localhost.

Deploy the app on cloud run

• Create a document AI processor , in this case , it is an OCR type processor.
• Create a .env file with the following environment variables:

  PROJECT_ID='YOUR_PROJECT_ID'
  LOCATION='eu'
  PROCESSOR_ID ='YOUR_DOCUMENT_AI_PROCESSOR_ID'

• Create a Dockerfile file
• Build the image and push it to conatiner registry

  gcloud builds submit --tag gcr.io/YOUR_PROJECT_ID/DEMO_NAME_OF_YOUR_CHOICE --timeout=1h

• Create a service in cloud run based on the built image.
• Access the app via the link provided by the cloud run service.