Hello! I am a Ph.D. candidate in the Uncertainty Quantification Group at the University of Houston, advised by Dr. Ruda Zhang.
My research develops trustworthy scientific AI for engineering and computational science, with a focus on uncertainty-aware scientific foundation models, model calibration, Bayesian optimization, and scientific discovery automation. I build methods that make scientific models more reliable, better calibrated, and more useful for engineering and scientific decision-making.
My earlier work focused on structural health monitoring, probabilistic damage detection under environmental variability, and model-form uncertainty in computational mechanics. That foundation now informs my current direction in trustworthy scientific AI, including uncertainty-aware scientific foundation models and agentic scientific ML workflows.
Research Themes
- Trustworthy Scientific AI: uncertainty-aware scientific foundation models, calibrated stochastic inference, and reliable decision-making.
- Agentic AI for Scientific Discovery: coding-agent workflows, reusable skills, evaluation harnesses, and scientific experimentation automation.
- Uncertainty Quantification and Model Error: stochastic subspaces, reduced-order modeling, Bayesian calibration, and computational mechanics.
More detail on papers and technical writing is available on the Publications and Blog pages.
Education
Ph.D. in Civil Engineering, University of Houston, May 2027*
Thesis: Quantifying and Reducing Model Uncertainty using Stochastic RepresentationsM.Tech (Research) in Civil Engineering, Indian Institute of Science, Bangalore, June 2023
Thesis: Structural Health Monitoring Accounting for Thermal Variability and Damage Using Approximate Bayesian Computation (ABC)B.Tech in Civil Engineering, Indian Institute of Technology, Roorkee, May 2018
Thesis: Design of Hydro Power Project
News
April 21, 2026 – Released the arXiv preprint Calibrating Scientific Foundation Models with Inference-Time Stochastic Attention.
March 20, 2026 – Received the Student Travel Award for the 17th World Congress on Computational Mechanics and 10th European Congress on Computational Methods in Applied Sciences and Engineering (WCCM/ECCOMAS 2026) by the United States Association for Computational Mechanics (USACM).
December 25, 2025 – Attended The Thirty-Ninth Annual Conference on Neural Information Processing Systems (NeurIPS 2025) at San Diego, California.
Contact
ayadav4 ‘at’ uh ‘dot’ edu
I am interested in collaborations on trustworthy scientific AI, agentic AI for scientific discovery, and uncertainty-aware foundation models. Please feel free to reach out by email. You can also reach me via LinkedIn.