About Me

I am a computer scientist in the Mathematics and Computer Science Division at Argonne National Laboratory, where I develop next-generation foundation models and intelligent AI systems to accelerate scientific discovery at scale. My research lies at the intersection of agentic AI, multimodal reasoning, trustworthy machine learning, and high-performance computing, with the goal of enabling AI systems that can reason over complex scientific data, collaborate autonomously, and operate reliably in scientific environments.

I lead and contribute interdisciplinary research across multiple scientific domains, including infrastructure resilience, intelligent transportation systems, smart mobility, HPC networks, hazard adaptation, and scientific visualization though seval DOE projects. My broader vision is to build reliable, safety-aligned, and scientifically grounded AI systems that accelerate discovery in complex, high-stakes scientific settings.

Prior to joining Argonne, I served as a Senior Data Scientist at General Electric, where I applied advanced machine learning to large-scale industrial systems. I earned my Ph.D. in Computer Science from the Indian Institute of Technology (IIT) Kharagpur, India.


My current research interest

  • Agentic systems: Our research investigates how autonomous AI agents can plan, reason, and coordinate in complex, dynamic environments, with a focus on designing scalable orchestration and workflow intelligence that enable reliable and efficient multi-agent decision-making.

  • Multimodal Reasoning for Scientific Discovery: We examines how AI systems can integrate and reason across heterogeneous data modalities, including text, images, time series and spatiotemporal information. We explore representation learning, cross-modal alignment, and compositional reasoning, along with the development of rigorous benchmarks to evaluate multimodal scientific understanding.

  • Foundation Models training: We studies how foundation models can be trained and adapted for scientific domains while preserving generalization, reasoning, and scalability, with an emphasis on data-efficient and domain-aware learning for reliable scientific use.

  • Reliable AI systems: We studies methods for ensuring that AI systems are robust, interpretable, and dependable in high-stakes environments. We focus on uncertainty quantification, hallucination detection, failure prediction, and risk-aware reasoning to support safe and predictable deployment in real-world scientific applications.

  • Science application areas: We also investigates how advanced AI methods can be applied to domains such as infrastructure, and high-performance computing systems to enable scalable, data-driven scientific discovery and resilient decision-making in complex environments.

Updates

  • Contributing to the ModCon Core Agentic Framework (CAF) team as part of the Genessis mission, started in FY26.
  • Leading the NOAA NIDIS AI Drought Assistant MCS team and the ESURE LDRD project, both started in FY26.
  • Contributing to RAPIDS3, SBIR digital Twins project, started FY26.
  • "LUMINA: Detecting Hallucinations in RAG System with Context-Knowledge Signals" accepted to ICLR 2026.
  • Panelist at ATARC’s Virtual Roundtable on Securing Innovation: Modernizing National Lab Infrastructure for the AI Era, held on January 15, 2026.
  • "MARSHA: multi-agent RAG system for hazard adaptation" published in Nature Climate Action.
  • "PGT-I: Scaling Spatiotemporal GNNs with Memory-Efficient Distributed Training" published in SC25.

Projects

  • (Active) The Transformational AI Models Consortium (ModCon)
  • (Active) Energy and National Security and Reliability (ENSURE) Foundation Model
  • (Active) Phase II: Composable Digital Twins for Science Network Infrastructures using Parallel Discrete Event Simulation.
  • (Active) The RAPIDS3 Institute for Artificial Intelligence, Computer Science, and Data
  • (Active) Enhancing ClimRR Capabilities to Better Support Electric Utility Applications and Technical Assistance
  • (Active) AuroraGPT: Large Language Model for Science and Engineering
  • (Active) Tachyon: Intelligent Multi-Scale Modeling of Distributed Resilient Infrastructure and Workflows for Data Intensive HEP Analyses
  • (Finished) Climate Action Through Large Language Models (CALLM): A Tool for Analyzing Climate Risk and Building Resilience
  • (Finished) RAPIDS2:A SciDAC Institute for Computer Science, Data, and Artificial Intelligence
  • (Finished) Composable Digital Twins for Science Network Infrastructures using Parallel Discrete Event Simulation
  • (Finished) Developing Machine Learning Surrogate Models For Regional Scale Transportation Network Models
  • (Finished) High-Performance Computing and Big Data Solutions for Mobility Design and Planning
  • (Finished) Community and Infrastructure Adaptation to Climate Change: An AI-Driven Research Tool
  • (Finished) A Demonstration Project on the Application of Artificial Intelligence for Identifying Hybrid Threats Affecting National Critical Functions in the United States and Allied Nations
  • (Finished) Accelerating graph-convolution-based deep learning framework for large-scale highway traffic forecasting with SambaNova

Honors and Awards

  • Received the Impact Argonne Award for enhancement of Argonne's reputation, March 2024.
  • Received second place in the NeurIPS Forecast TracK: CityLearn Challenge 2023 as Team Vanguards in collaboration with Yangxinyu Xie and Ngoc Tran. The recognition includes a cash prize and co-authorship for a summary manuscript, December 2023.
  • Received Impact Argonne Award in recognition of the outstanding contributions to the AI for Science training series and enhancement of Argonne's reputation by reaching out to its next generation of scientific users, January 2022.
  • Travel award: 14th Women in Machine Learning (WiML) Workshop co-located with NeurIPS, Vancouver, Canada December, 2019
  • Selected to participate in the 2nd Heidelberg Laureate Forum as one of the 100 most qualified Young Researchers in Computer Science to meet a number of Turing Awardees, September, 2014
  • The paper "Characterization of Noise in Kinetic Depth Images: A Review" as one of the 25 most downloaded papers for April and May 2014

  • One of the 7 finalist of Samsung Innovation Award 2014. In total 50 groups had participated for the award. -- News Article

  • Received TCS research fellowship during Ph.D.

  • Travel award from DST, Govt. of India for VISAPP 2015, and from Rajaghari Fund of IIT Kharagpur for 2nd Heidelberg Laureate Forum

  • Received GATE (Graduate Aptitude Test in Engineering) scholarship duting masters

Outreach