Nirmit is a researcher working at the intersection of materials science and autonomous experimentation. His current research develops algorithmic decision layers for closed-loop experimentation for materials discovery. He is working on implementing physics-informed priors over search spaces, multi-fidelity learning, agentic hardware orchestration, and physical AI tools. More broadly, he is interested in how scientific foundational models can serve as prior-generating layers for combined systems. Nirmit entered this nexus with a foundation in electrochemistry and battery materials, spanning glovebox Lithium metal cell fabrication, high-throughput electrolyte screening, and physics-based degradation modelling. Before LLNL, he worked on experimental-computational pipelines at the German Aerospace Center (DLR) and performance modeling of batteries at Carnegie Mellon University (CMU). 

Research Interests: 

• Multifidelity optimization (Bayesian optimization, evolutionary search)
• Agentic infrastructure for sim-to-real integration
• Autonomous experimentation and self-driving labs
• Scientific foundation models
• Electrochemical energy storage systems

MS, Energy Science, Carnegie Mellon University, Pittsburgh, Pennsylvania

BE, Electrical and Electronics Engineering, BITS Pilani, India

For a full list: link to Google Scholar