This post contains an exercise to predict customer churn, a typical Machine Learning use case. I used GCP services.

This demo shows:

1. How to easily implement an existing model already available in BigQuery UI (BigQuery ML).
2. Show reporting capabilities. Connect a BI tool (Looker, Qlik Sense, Power BI, etc.) to the BigQuery model.
3. Explain how to connect a Jupyter notebook directly to BigQuery
4. Depending on the audience, you could use this demo to start the conversation on doing MLOps.

The demo uses the Telco Customer Churn from Kaggle.

You have the materials for the demo here.