Define-ML provides a structured framework for addressing the unique challenges of ideating ML-enabled systems. By extending Lean Inception with targeted activities—Data Source Mapping, Feature-to-Data Source Mapping, and ML Mapping—the approach helps teams ground ideation in real data constraints, clarify ML feasibility, and align proposed features with business objectives.
The approach is being developed in the context of the Ph.D. of Silvio Alonso at PUC-Rio (Brazil). It has evolved after a series of empirical studies conducted at the ExACTa PUC-Rio lab and is currently being applied in practice by partner companies.
The Miro board provides an example based on a toy problem and an empty board to be filled in the context of a Define-ML workshop. A research paper describing Define-ML and its initial evaluation has been submitted to the EuroMicro SEAA conference, and will be made available in case of acceptance.
