Research software · scientific computing · Bayesian inference
MSc Scientific Computing with Data Science, University of Bristol (2026)
I build scientific software that people can trust: tested, documented, and honest about where it breaks.
- kappa-kinisi · MSc dissertation. Found why a small fraction of kinisi diffusion fits go wrong (14 of 64,000), traced each one to a single eigendirection of the covariance matrix, and designed an adaptive eigenvalue floor that fixes them without changing healthy fits.
- qens · Bayesian analysis of quasi-elastic neutron scattering data from ISIS
.nxspefiles: joint fit across all Q-bins, MCMC with convergence checks, and a plug-in model registry. Docs - nxcat · A small metadata catalogue for neutron data: NeXus ingest, FastAPI, React and TypeScript, PostgreSQL search, and a plain-English search where the AI can only suggest a filter that the user confirms.
- scisolve · A numerical solver where every number in an answer must come from code run in the same session, and every answer names the check it passed.
- arxiv-rag · A retrieval system over research papers, where every answer quotes its source or declines.
Python · NumPy · SciPy · emcee · FastAPI · React/TypeScript · PostgreSQL · pytest · GitHub Actions · Docker · Read the Docs
- MSc Scientific Computing with Data Science, University of Bristol, 2026 (SARTHI Foreign Scholarship)
- BE Mechanical Engineering, Savitribai Phule Pune University, 2020
- Previously in ML research (Spartifical) and engineering roles in industry
Open to research software and scientific computing roles. Get in touch.