As a Machine Learning engineer, when I need to work on a new problem, I start with creating a Jupyter notebook and start coding, but after some iteration, the project structure starts looking like this. Each notebook might contain a different model configuration along with metrics. Doing experiments this way…