Maziar Raissi

Associate Professor of Applied Mathematics
University of California, Riverside

Email Google Scholar GitHub YouTube LinkedIn ORCID

I build AI that respects the laws of physics.

Most real problems in science and engineering come with a few noisy measurements and a lot of known physics. My group builds neural networks that learn from data while obeying the equations behind fluids, materials, energy systems, and the human body.

I created Physics-Informed Neural Networks (PINNs), now a standard tool in scientific machine learning. The original paper has been cited more than 33,000 times (Google Scholar) and received the 2024 ICBS Frontiers of Science Award.

Prospective students and postdocs

I am recruiting PhD students and postdocs in scientific machine learning: physics-informed neural networks, AI agents for scientific simulation, and digital twins of physical systems.

Email me with the subject line "Prospective student" or "Prospective postdoc", your CV, and two or three sentences on a problem you would like to work on. PhD applicants should also apply to the UCR Mathematics PhD program.

Research now

AI agents for scientific simulation. AI Lead on MAESTRO, a DOE Genesis Mission project on multi-agent AI for subsurface energy.
Physics-informed models for power systems, subsurface energy, and cardiovascular blood flow.
Self-learning models for robotic remanufacturing (NSF).

Supported by DOE, NSF, the University of California, and Procter & Gamble.

Selected work

Physics-informed neural networksJournal of Computational Physics, 2019 · code · US Patent 10,963,540

Hidden fluid mechanics: learning velocity and pressure fields from flow visualizationsScience, 2020 · code

SimNet, now NVIDIA PhysicsNeMoI started SimNet at NVIDIA: GPU software that simulates physics with AI · SC19 keynote demo

Earlier projectsDeep hidden physics models, numerical Gaussian processes, multi-fidelity modeling, and more

Full list: Google Scholar.

Teaching

Applied Deep Learning, a free course with over 1 million views on YouTube and 3,500+ stars on GitHub. Short tutorials: deep learning, Gaussian processes, backpropagation.

Path

UC RiversideAssociate Professor of Applied Mathematics (2025–), Assistant Professor (2023–2025)

University of Colorado BoulderAssistant Professor of Applied Mathematics (2020–2023)

NVIDIASenior Software Engineer (2018–2020)

Brown UniversityAssistant Professor of Applied Mathematics (Research) and postdoc, Division of Applied Mathematics

University of Maryland, College ParkPhD, Applied Mathematics & Statistics, and Scientific Computation

George Mason UniversityPhD, Mathematics

Recognition

ICBS Frontiers of Science Award, 2024 (for the PINNs paper)
Top-cited scientists, Stanford–Elsevier database (2026): #5 of 23,982 in Applied Mathematics (data)