Discovery and development in chemical, biological, and materials sciences remains highly dependent on costly and time-consuming experimental programmes. Conventional statistics-based Design of Experiments (DOE) methods have made a substantial contribution to lessening this burden, reflected in the widespread use of respected software such as JMP and Minitab. But machine learning offers the opportunity to go a step further, particularly for multi-component problems, systems with non-linear effects, or to enable adaptive DOE in which an experimental programme is refined as it progresses. In this webinar we discuss, with case studies and a software demonstration, the use of the Alchemite™ machine learning software for DOE. Alchemite™ typically results in a 50-80% reduction in the number of experiments required compared to conventional DOE methods.