Project Overview

The project began as a single Jupyter notebook used for experimentation. Once the modeling approach was validated, the notebook was refactored into a production-style pipeline composed of independent Python modules.

Key ideas demonstrated:

• Separation of concerns

• Script-based execution instead of notebooks

• Deterministic, repeatable pipelines using DVC

• Parameter-driven experimentation

Tools used:

• Python

• DVC

• Git & GitHub

• Cookiecutter Data Science