Emerging Approaches

Given the lack of established data science-specific project management methodologies, organizations have started to formalize their own data science project management approaches that typically apply agile concepts onto CRISP-DM. Like open source data science projects and algorithms, efforts to publish data science project management approaches will help the data science field mature and benefit the entire community. Conceptually, the combination of data science-specific processes and agile project management make these attractive for data science project management approaches.

The Microsoft Team Data Science Process is a comprehensive methodology defines roles, processes, and templates that are largely inspired from CRISP-DM and Scrum.

Domino Data Lab’s Data Science Lifecycle likewise combines elements of CRISP-DM and general agile approaches. Presented as a “best practices” guide, it is not as defined as TDSP.