I am a Research Associate at the Indian Institute of Science (IISc), Bangalore. I am currently contributing to the development of India's first Scientific Foundation Model. I obtained my Master of Science in Mathematics from the National Institute of Technology, Kurukshetra (2023-2025), mentored by Prof. A.S.V. Ravi Kanth. Previously, I obtained a B.S. in Mathematics (Honours) at the University of Delhi (2020-2023). I am actively looking for a PhD position.

My long-term goal is to bridge the gap between theoretical machine learning and AI for Industry, deploying robust methods for scientific computing and discovery where AI faces complex real-world challenges. Specifically, I pioneer the application of physics-informed neural networks (PINNs), operator learning, scientific machine learning (SciML), and scientific foundation models to solve partial differential equations (PDEs) arising in industrial engineering. By building physics-aware architectures capable of learning from sparse data, these models directly accelerate industrial pipelines, with applications spanning from predicting magnetic fields in large-scale stator geometries to designing robust foundation models that safely generalize across industrial PDEs.

Research Interests: AI for Science; Foundation Models; Neurosymbolic AI; Scientific Machine Learning; Physics-Informed Neural Networks; Operator Learning; Meta-Learned Universal Solvers; Latent Representations Structure.

If any of my research excites you, or if you just want to talk about blending deep learning with the physical sciences, I am always open for collaboration and chats. When I step away from research, you can find me cycling, cooking, sketching, or experimenting with novel ways to visualize complex mathematical concepts through generative diagrams.

Mamta Saini