/ml-cloud-unstructured-data

Primary LanguageJavaScript

Serverless Functions

Code

Download the credentials.json file of your service account from gcloud.

Use the function.sh script to run the funcions locally (you'll need Docker for that)

./functions.sh ocr develop

./functions.sh ocr-results develop

./functions.sh nlp develop

Invoke Locally

OCR

curl --header "Content-Type: application/json" \
  --request POST \
  --data '{"data": {"bucket": "k-webinar-spainai-2020-input", "name": "LIDL-20200102-20200108.pdf"}}' \
  http://localhost:8090

OCR RESULTS

curl --header "Content-Type: application/json" \
  --request POST \
  --data '{"data": {"bucket": "k-webinar-spainai-2020-output", "name": "results/ocr/LIDL-20200102-20200108/output-1-to-20.json"}}' \
  http://localhost:8090

NLP

curl --header "Content-Type: application/json" \
  --request POST \
  --data '{"data": {"bucket": "k-webinar-spainai-2020-input-reviews", "name": "Oikos-Caramelo_r_932849.txt"}}' \
  http://localhost:8090

Cloud Resources

Cloud Repositories

  • Three repositories

Cloud Build

  • Three Cloud Build Triggers

Cloud Storage

  • Three Cloud Storage Buckets

BigQuery

  • A couple of datasets with two tables each

Cloud Functions

  • Three Funcions

Cloud Vision and NLP

  • API Requests