Artificial intelligence is gaining prominence in diverse industrial sectors, transforming procedures and decision-making with advanced functionalities that were previously unattainable or not easily achievable with traditional software. Industries are allocating resources towards artificial intelligence-driven solutions, which have the potential to enhance operational efficiency and foster innovation within their respective domains. Regrettably, machine learning also presents novel obstacles, such as managing uncertainties arising from the decision-making process. This paper shows the design and implementation of a framework that allows modeling and simulating time-based software and hardware systems to assess the potential impacts of uncertainties propagating across the system in various operational environments. This is crucial as those uncertainties could result in costly system breakdowns and influence other non-functional prerequisites like system efficiency, dependability, and even security. Our methodology involves establishing a novel domain-specific language with formal execution semantics delineated using stochastic colored Petri nets. The resultant models can subsequently undergo static analysis and simulation via an automated conversion to stochastic colored Petri nets, facilitating the computation of diverse reliability and efficiency metrics.