Lead Research Software Engineer - Huang Lab, Icahn School of Medicine - New York
(2026-03)
Single-cell GRN software | Open source
- Built and released RustScenic, a Rust/Python implementation of SCENIC/SCENIC+ compute workflows, replacing fragile Dask/Java/conda-heavy stages with faster, lower-memory, deterministic, cross-platform software.
- Work directly with Huang Lab researchers on real multiome validation, turning weekly lab feedback, test runs, and external tester reports into benchmark artefacts, memory/runtime fixes, and release priorities.
- Shipped v0.4.5 to PyPI with Linux/macOS/Windows wheels, live docs, typed Python support, Zenodo DOI, MIT licence, branch-protected CI, and reproducibility artefacts.
- Benchmarked against pySCENIC, arboreto, ctxcore, and public multiome datasets: AUCell 27x faster (0.25s vs 6.81s, r = 0.984), cisTarget r = 1.0000, and 100k-cell peak RSS 6.3 GB vs >40 GB legacy stack.
Volunteer Research Assistant - Grieshop Lab, University of East Anglia - Norwich, UK
(2026-03)
- Run Nextflow population-genomics workflows on the Hali HPC cluster, supporting reproducible pipeline execution and large-scale genomic data analysis.