Example journey. This page shows how we run this type of migration. It is not a named client result. References from comparable projects are available on request.
This example journey shows how a telecom operator would bring BigQuery, Dataproc and off-platform churn models together on Databricks on Google Cloud: one governed lakehouse for customer, usage and network data, with churn models trained and served on the same data.
| Type | Example migration journey |
|---|---|
| Industry | Telecom |
| Migration | Google Cloud |
| Technologies | Databricks on Google Cloud, MLflow, Model Serving, Unity Catalog |
Starting point
A telecom operator stores customer data in BigQuery and processes network events on Dataproc. Data scientists export data to build churn models elsewhere, so models drift away from production data.
How we would approach it
- Run on Google Cloud: deploy Databricks on Google Cloud close to existing storage.
- Unify: bring customer, usage and network data into one medallion lakehouse.
- Convert: migrate BigQuery SQL and Dataproc jobs, and validate key reports.
- Operationalize ML: build churn models with MLflow on the governed tables and serve scores to retention teams.
Value the operator should expect
- One platform for network analytics, BI and ML.
- Models trained and served on the same governed data.
- A base for revenue assurance and care assistants.
What to measure
Data platform cost, model refresh time, churn campaign response, and time to add a new data source.
Read the full BigQuery and Dataproc to Databricks on Google Cloud guide →
