A series of articles that takes you through the concepts used in KMDS - it is a 10000 ft view, but it gives you the ideas used:

1. AI view versus data science view to building intelligent solutions: AI vs Data Science
2. Going from a business use case to a data science task view of the data: kmds-dataset-util
3. Going from a task view to a dataset that is ready for modeling and featurization: dd-parser-cleaner
4. Going from a modeling/featurization ready dataset to features and models feautrization and modeling