customer-churn-example

Welcome to Continual's customer churn example. For a full walk through, please visit the documentation. You can also find some great context at this blog post as well.

Note: This project is designed to run with Snowflake. You should be able to adapt it to other warehouse vendors pretty easily, but let us know if you have any issues.

Running the example

Download the source data at Kaggle:

kaggle competitions download -c kkbox-churn-prediction-challenge

You can then upload it to your cloud data warehouse of choice in the manner of choice. For convenience, we've included a few short scripts to upload this manually in snowflake.

1. First run the ddl.sql file

2. Then you can upload the CSVs downloaded from kaggle via snowsql. Copy the commands in snowsql_staging.sql into snowsql and execute them.

For dbt users

If you're using dbt, you'll now just be able to use the dbt project provided. dbt_project.yml is configured to use the continual profile. You'll either need to change it or create your own in your profiles.yml file. Then you can execute:

dbt run

To build all the required tables/views.

When dbt is finished, you can then run

continual run ./continual/featuresets ./continual/models

You'll of course need to ensure that you have an account in continual and have logged in with the CLI and have set up a default project.

You're now done! You can check out the Continual Web UI to monitor the model and see results as it finishes. (Note: It should take about 2 hrs to finish.)

For non-dbt users.

We highly recommend using dbt for your transformations. If this is not feasible, we've provided to sql scripts feature_engineering.sql and prediction_engineering.sql that you can run in Snowflake to build out all required tables/views.

After that, you can simply execute the following:

continual push ./continual/featuresets ./continual/models

You'll of course need to ensure that you have an account in continual and have logged in with the CLI and have set up a default project. Note: If you modify any of the table names in the .sql scripts, you'll want to update the queries in the .yaml in continual/featuresets and continual/models accordingly.

You're now done! You can check out the Continual Web UI to monitor the model and see results as it finishes. (Note: It should take about 2 hrs to finish.)

Having issues?

Feel free to let us know if you have any issues, or you can open a PR with any suggested modifications.

Thanks!