I will develop custom physics informed neural networks pinns in pytorch

Are your FEM/CFD simulations too slow or computationally heavy?

I am a PhD computational researcher and Visiting Researcher at Imperial College London, specializing in Scientific Machine Learning (SciML).

I build custom Physics-Informed Neural Networks (PINNs) in pure PyTorch from scratch, bypassing high-level wrappers to guarantee maximum control, debuggability, and performance for your physics constraints.

What I can execute for you:

• Solving complex forward and inverse PDEs, including extracting unknown parameters from experimental data.

• Creating fast, ML-based surrogate replacements for computationally expensive traditional simulations.

• Debugging and accelerating your existing PyTorch or SciML codebases.

Recent Executions:

• Modeled high-Re turbulent flows (Re=10,000 Navier-Stokes).

• Solved complex elastoplasticity problem via Hertz Contact PINNs.

Why Hire Me? You aren't just getting a coder; you are getting a strategic computational partner. I ensure your models are mathematically flawless.

Please message me with your PDE and scope before ordering!