We present progress on our project aimed at enhancing predictive simulations in mechan- ics through the integration of Learning the effective dynamics (LED) [1] and Bayesian methods. Our framework, built on Korali [1] — an open-source platform for optimization and uncertainty quantification—encompasses a multiphysics microscale engine and an adaptive micro-scale surrogate framework based on LED. This comprehensive framework encompasses a multiphysics microscale engine and an adaptive surrogate framework, augmented by hierarchical Bayesian methods, ensuring iterative enhancements in simulation precision and efficiency. Our ramework’s application in patient-specific cancer immunotherapy illustrates its po- tential in revolutionizing personalized treatment strategies. This demonstration emphasizes the framework’s practical applications in bioengineering and medical sciences, par- ticularly in tailoring patient-specific therapies. Our presentation showcases the tangible impact of this integrated approach on predictive simulations and personalized medicine. This work is part of DCoMEX [3], a European High Performance Computing Joint Un- dertaking project. REFERENCES [1] Vlachas, P. R., Arampatzis, G., Uhler, C., Koumoutsakos, P. (2022). Multiscale simulations of complex systems by learning their effective dynamics. Nature Machine Intelligence, 4(4), 359-366. [2] Martin, S.M., W ̈alchli, D., Arampatzis, G., Economides, A.E., Karnakov, P. and Koumoutsakos, P., 2022. Korali: Efficient and scalable software framework for Bayesian uncertainty quantification and stochastic optimization. Computer Meth- ods in Applied Mechanics and Engineering, 389, p.114264 [3] DCoMEX: http://mgroup.ntua.gr/dcomex