BRAID aims to research new methods for processing data from next-generation science experiments efficiently and reliably in continuum computing environments, while satisfying requirements for rapid response, high reconstruction fidelity, data enhancement, data preservation, model training, etc. It will also benefit scientific projects at DOE facilities by enabling them to advance from today’s, largely manual, collect now, analyze later data analysis paradigm to a new automated, AI-enhanced, self-documenting collect, analyze, feedback, document paradigm. Finally, it will contribute to understanding of how such research automation can be used to improve the quality of generated data.