I'm about to start with a ML learning time-series forecasting project which includes large-quantities of (2-million) time-series stored in JSON-files.
The first step will involve feature engineering. Then I would like to experiment with different models: probably, DL, Comb & Hybrid. Ultimately I would like to visualize results on top of, eg,Flask. I prefer to do most of it in Python.
Wat set-up/project structure would be recommended for such project? Is it recommended to store raw data in a database for example? What is recommended in terms of efficiency?