The integration of novel autonomous ocean monitoring and probabilistic ocean forecasting can be most beneficial to the Gulf of Mexico (GoM) and its communities and stakeholders. Such integration combines systems for high-resolution stochastic ocean modeling, multi-platform autonomous observing, data assimilation, path planning, adaptive sampling, real-time operations, and of course human interactions. The first steps towards this integration for the GoM were demonstrated in real-time during the collaborative Mini-Adaptive Sampling Test Run (MASTR) ocean experiment. The main MASTR effort occurred from February to April 2024, as part of the collaborative “Understanding Gulf Ocean Systems (UGOS-3)” initiative sponsored by the Gulf Research Program of the U.S. National Academies of Sciences, Engineering, and Medicine. The emphasis of the present contribution is on some of the real-time MASTR results. They include large-ensemble forecasting of physical ocean fields, uncertainties, and risks, predicting reachable regions and optimal paths for aircraft and marine platforms, forecasting optimal ocean sampling times and locations, and evaluating the skill of forecasts by comparison with observations.