A multi-national manufacturing organisation.
The Business Challenge
In order to do stock optimisation simulation this organisations data needs to be transformed and validated in a series of complex steps.
DataIKU was already being used as data science platform to wrangle the data into the shape needed for the eventual solution. It was discovered that a repetitive validation and selection process was being followed for each of the various input sources, but that these processes varied in quality and steps, based on each persons style and skills. The solution had to address the inconsistencies that resulted. Applying Data Engineering techniques, a few processing patterns were designed, and implemented as DataIKU plugins and macros. Using the DataIKU public API, this now provides the client with a few standardised options to automatically apply standard validation and selection processed.
The Value Proposition
Using the DataIKU API, these repetitive steps could be automatically generated, which reduced the validation and selection development time by an order of magnitude, while also automatically taking care of naming standards and best practises across all validation processes.
No posts were found.