Explorative Questions:

[BradKML]

1. Given a model with singular function, of pre-defined inputs (assuming monovariate prediction), can a flask model be generated without excess inputQueryXXX = request.form['queryXXX']? Can this be considered a form of D.R.Y. design?
2. What is the debate about when it comes to "pickling models (and tabular data) are bad" other than arbitrary code execution? What are the most fitting alternative?

• Regarding claimed alternatives, there are JobLib, Keras, and PMML/ONNX https://towardsdatascience.com/5-different-ways-to-save-your-machine-learning-model-b7996489d433
• Regarding risks https://faoel.medium.com/stop-using-pickle-to-save-machine-learning-models-eaa46e8e561a https://uwekorn.com/2021/04/26/implications-of-pickling-ml-models.html

3. Regarding the code itself, what is CustomerID (cannot be found in the repo search), how is it tied to the tenure_group as annual segments?