Nesar Ramachandra, PhD

I am a Computational Scientist at Argonne National Laboratory, Chicago working at the intersection of artificial intelligence and scientific research. My work focuses on developing innovative computational methods for scientific analyses, particularly in the fields of cosmology and astrophysics.

Current Position

At the Computational Sciences division at Argonne, I develop and apply machine learning techniques to address complex scientific challenges. My work involves:

• Developing uncertainty quantification methods for scientific models
• Designing emulators for computationally expensive simulations
• Implementing multi-modal foundation models on heteregenous scientific datasets
• Developing Agent-led workflows for scientific problem solving.
• Implementing Deep Learning pipelines for cluster cosmology, strong lensing analysis
• Creating realistic synthetic extragalactic catalogs from exascale simulations.

Research Interests

• Machine Learning for Science: Developing AI methods that incorporate physical constraints and domain knowledge
• Cosmological Structure Formation: Understanding the formation and evolution of the cosmic web
• Uncertainty Quantification: Creating probabilistic models that provide reliable uncertainty estimates
• Dark Matter Physics: Investigating the properties and distribution of dark matter in the universe

Background

Before joining Argonne as a research scientist, I was a Postdoctoral Fellow at the High Energy Physics Division at Argonne and a KICP Associate fellow at the University of Chicago.

I completed my PhD in Physics and Astronomy at the University of Kansas under the supervision of Professor Sergei Shandarin, focusing on computational and physical cosmology.

My academic journey began at BITS Pilani, India, where I earned my Integrated Masters in Physics with an honors thesis from the Indian Institute of Astrophysics, Bangalore.