Living systems consume energy to coordinate biomolecular assembly in space and time, tuning their structure and dynamics with a precision unmatched by human-engineered systems. Embedding passive building blocks in active fluids is a promising avenue for realizing similarly life-like assemblies with synthetic materials. However, the wide spectrum of relevant length and time scales makes this task challenging, and existing approaches to simulate active materials are limited to small system sizes and/or short times. We introduce a new computational method that models soft materials advected by an active fluid as particles moving in a spatiotemporally correlated noise field. This technique is simple, efficient, and interfaces with the popular HOOMD-blue simulation package, enabling large-scale simulations. Importantly, noise parameters can be inferred from experimental velocimetry measurements, allowing the model to be tailored to specific active materials. To demonstrate the method, we investigate the activity-driven organization of colloidal particles. The particles assemble into a range of dynamic structures, from rotating droplets to system-spanning, flowing networks. Notably, the sizes and dynamics of these structures can be tuned via the active noise correlation length and time. By elucidating how active flows drive nonequilibrium assembly, our results provide insights into the design of life-like materials. Moreover, our technique can be used to simulate a wide array of active fluids and other systems with spatiotemporally correlated noise.