I'm a computational physicist finishing my PhD at the University of Rhode Island. I build fast surrogate models for black hole collisions by day, and tinker with ML for finance problems by night.
I like making things that are fast, reliable, and actually useful.
- BHPTNRSurrogate — Surrogate model for gravitational waveforms. 10⁶× faster than numerical simulations, used by multiple research groups.
- AML-detection-using-AI — Temporal GNN for anti-money laundering detection. 0.84 PR-AUC on 9.5M transactions, outperformed XGBoost by 41%.
- SEOBNRv4CE — Waveform model with built-in uncertainty quantification using Gaussian Process Regression.
Python (7+ years), PyTorch, SQL, Bayesian inference, time-series modeling, HPC (MPI, Slurm, GPU acceleration), and building software that ships with tests, CI/CD, and docs.
Data science and ML roles in finance, quantitative analytics, and research software engineering.