Akash Srivastava is a Research Scientist and Principal Investigator (PI) at the MIT-IBM Watson AI Lab in Cambridge, MA, and Chief Architect of Large Language Model Alignment at IBM Research. His work focuses on generative modeling , Bayesian inference , and machine learning for constrained engineering design . He previously conducted PhD research at the University of Edinburgh under Dr. Charles Sutton and Dr. Michael U. Gutmann on variational inference for generative models using deep learning. His research spans Neuro-Symbolic AI , Language Model Alignment , and Synthetic Data Generation , with applications in 3D modeling , urban logistics , and material science . Recent publications highlight advancements in diffusion models , continual learning , and privacy-preserving data synthesis . As a PI, he collaborates with MIT faculty like Prof. Faez Ahmed and Prof. Rafael Gomez-Bombarelli on projects such as generative modeling for mechanical systems , synthetic data in decision-making , and greener delivery networks . He has received funding through a DARPA grant for machine common sense research.
