Building scalable Python/C++ systems for distributed experimentation, model evaluation, and ML-ready infrastructure.
Focus | Experience | Projects | Skills | Analytics | Contact
- Distributed systems and high-performance computing workflows across large CPU/GPU clusters.
- AI/ML infrastructure for training, evaluation, and large-scale experiment orchestration.
- End-to-end Python/C++ platform development for compute-intensive scientific and ML workloads.
- Data ingestion, preprocessing, post-processing, and validation pipelines at scale.
- Research Engineer / Applied Scientist, SimCenter and DesignSafe (NSF) (2023-Present): Designed distributed Python/C++ frameworks on TACC systems, built high-throughput execution pipelines, and optimized MPI workloads across thousands of CPU cores.
- Graduate Research Assistant / Founder and Lead Developer, University of Washington (2022-Present): Created and now own Femora, conceived the original architecture, and authored the core framework end-to-end for automated model generation and distributed experimentation.
| Project | Role | What It Delivers |
|---|---|---|
| Femora | Owner, original creator, and primary author of GeotechUW/Femora | Automated model generation, scalable preprocessing, and distributed experiment execution. |
| FemoraX: JAX GPU FEM Solver | Architect and core developer | Optimized finite element solver in JAX for GPU acceleration, fast kernels, and nonlinear analysis workflows. |
| SimCenter EE-UQ | Contributor | Earthquake engineering and uncertainty quantification workflows with scalable simulation tooling. |
| OpenSees | Contributor | High-performance structural simulation ecosystem for nonlinear finite element analysis. |
- ML and AI Infrastructure: PyTorch, JAX, Scikit-learn, experiment design, training and evaluation workflows.
- Distributed Systems: MPI, OpenMP, high-throughput orchestration, cluster-scale execution.
- Languages: Python (expert), C++ (advanced), SQL, MATLAB.
- Data and Platform Tooling: NumPy, Pandas, HDF5, VTK/PyVista, Jupyter, Git, CI workflows, Linux, Docker.
If you are building distributed simulation systems, AI/ML infrastructure, or GPU-first numerical tooling, I would be glad to connect.
