I build physics-informed neural networks, neural operators, and SciML surrogates for simulation acceleration and PDE solving — work that sits at the intersection of deep learning and physical systems.
What's included:
Neural operator or physics-informed model for your specific problem
JAX or PyTorch implementation with full documentation
Performance benchmarks and results report
Clean, reproducible codebase with training scripts
Past work: Co-built AdS-CFT SciML Engine achieving >500× speedup over classical O(N³) PDE solvers at 3% L2 error using SIREN and FiLM-conditioned neural operators. Independent quantum circuit geometry research using Quantum Natural Gradient in PennyLane.
I build physics-informed neural networks, neural operators, and SciML surrogates for simulation acceleration and PDE solving — work that sits at the intersection of deep learning and physical systems.
What's included:
Neural operator or physics-informed model for your specific problem
JAX or PyTorch implementation with full documentation
Performance benchmarks and results report
Clean, reproducible codebase with training scripts
Past work: Co-built AdS-CFT SciML Engine achieving >500× speedup over classical O(N³) PDE solvers at 3% L2 error using SIREN and FiLM-conditioned neural operators. Independent quantum circuit geometry research using Quantum Natural Gradient in PennyLane.