ExpEngine: A Tool for Data Analytics Workflow Optimization Through User-Driven Experimentation

With the exponential growth of data analytics workflows, optimizing them has become crucial. In particular, there is a need to go beyond what the AutoML tools currently offer by considering different aspects of users (domain experts and data scientists) in the optimization process. Responding to this need, we propose a tool framework comprising an Experimentation Engine and a Domain-Specific Language. Together, they allow users to specify not only what needs to be optimized, but also how: in which steps, and with the desired degree of user involvement. The sources and demo video of our tool framework are available at https://github.com/ExtremeXP-VU/ExpEngine-ACSOS2024.