CV

Link to my Full CV - here

Education

• Ph.D in Computer Science, University of California, Irvine, 2026 (expected)
• Integrated M.Sc. (B.S.+ M.S.) in Mathematics, National Institute of Technology, Rourkela, 2016

Work experience

• Winter, 2023 - Present: Graduate Research Assistant
  • ETAD Lab, UCI
  • Duties : My research focuses on advancing fairness and interpretability in Machine Learning, particu- larly through fair representation learning and fair multimodal learning frameworks that are robust to input perturbations and come with theoretical performance guarantees. I have expertise with training deep neural networks, neuro-symbolic AI (e.g., Kolmogorov–Arnold Networks), and Large Language Models (LLMs), under fairness constraints.
  • Developed adaptive algorithm to mitigate algorithmic bias with convergence guarantees with adversarial robustness in KAN-based adversarial framework, achieving results that surpass SOTA adversarial models (e.g., ROAD) and address the fairness–accuracy trade-off without compromising predictive reliability.
  • Current work focus on developing fair Multimodal learning technique, with performance guaran- tees to handle distributional shifts under adversarial conditions, whilst also maintaining modality alignment. We propose Wasserstein DRO technique in an adversarial multimodal setup to tackle problems like bias amplification, modality misalignment and test-time distribution shifts

  • Second project focus - Designing an LLM-based Agentic AI framework for personalized student–project matching in capstone courses. Alongwith that the framework also aims to provide an explanation of the decision taken by the agent to enhance model prediction interpretability.

  • Supervisor: Prof. Sergio Gago-Masague
• Fall 2022 - Fall, 2023: Graduate Research Assistant
  • Dept. of Computer Science, UCI
  • Duties included: We worked on deep symbolic regression techniques using reinforcement learning algorithm which uses the policy gradient concept to produce mathematical expression(s) in the form of a syntax tree which is generated stochastically. The work focuses on improving the reward function through optimization technique by simultaneously improving the learning algorithm through noisy data inputs.
  • Supervisor: Prof. Eric Mjolness
• Fall 2019 - Present: Teaching Assistant
  • Dept. of Computer Science, UCI
  • Courses: Capstone Project (CS 180, CS 297P), System Design (ICS 53), Boolean Algebra and Discrete Mathematics (ICS 6B).

Skills

• Programming Language: C, C++, Python (Pandas, Scikit-learn, Matplotlib, Seaborn), SQL
• Deep Learning packages: PyTorch, TensorFlow
• Software: MATLAB, Anaconda, Jupyter Notebook, Mathematica, Visual Studio, PyCharm, GitHub, CUDA/MPS