Variation in patient response to glucagon-like peptide-1 receptor agonists (GLP-1 RAs) limits the full potential of these therapies for type 2 diabetes mellitus (T2DM) and obesity. This project addresses that gap by using large-scale clinical and genomic data to identify who benefits most from treatment and why. Machine learning methods will be applied to data from the All of Us Research Program, which includes longitudinal clinical records and genomic information from more than 600,000 participants, with strong representation of historically underrepresented populations. Genome-wide association analyses will first identify genetic variants linked to changes in glycemic control and weight. Those genetic signals will then be integrated with clinical factors to train supervised machine learning models that predict treatment response. Outcomes focus on clinically meaningful improvements in fasting glucose, HbA1c, weight, and body mass index, with model performance evaluated using standard predictive metrics. Results are expected to clarify the combined genetic and clinical drivers of GLP-1 RA effectiveness and to support more precise, data-informed treatment decisions for diabetes and obesity care.