Kushal Kolar
PHD Candidate
New York, New York, United States
Summary
🤩
Rockstar🎓
Top SchoolKushal Kolar is a PhD candidate and engineer with a decade of experience at the intersection of neuroscience and scientific software, currently developing GPU-accelerated visualization and numerical kernels for large-scale neuronal calcium and voltage imaging at NYU Tandon. He has a strong open-source track record—contributing serialization and model-saving features to tslearn for time-series ML and co-developing fastplotlib for high-dimensional GPU plotting—making complex analysis pipelines more reproducible and performant. Previously he built and released analysis tools for neural imaging and behavior at the Michael Sars Centre and contributed automated behavior-analysis tooling at EMBL. Kushal blends deep domain expertise in neurobiology with practical systems-level engineering across the Python data stack, and often optimizes low-level numerical code to unlock faster discovery workflows.
11 years of coding experience
5 years of employment as a software developer
Doctor of Philosophy - PhD Biomedical Engineering, Doctor of Philosophy - PhD Biomedical Engineering at NYU Tandon School of EngineeringBachelor of Science Integrated Science: Neurobiology & Biochemistry, Bachelor of Science Integrated Science: Neurobiology & Biochemistry at The University of British ColumbiaDoctor of Philosophy 1st 2 years of PhD, Doctor of Philosophy 1st 2 years of PhD at The University of North Carolina at Chapel Hill
Github Skills (13)
scikit-learn10
data-analysis10
serialization10
machine-learning10
time-series10
serializer10
python10
clustering10
scikit10
hdf9
kmeans9
json8
pickle8
Programming languages (9)
Github contributions (5)
tslearn-team/tslearn
Jan 2020 - Dec 2020
The machine learning toolkit for time series analysis in Python
Role in this project:
ML Engineer
ML EngineerContributions:22 commits, 4 PRs, 42 comments in 10 months
Contributions summary:Kushal primarily contributed to the serialization and model saving capabilities of the `tslearn` library, particularly for machine learning models related to time series analysis. They added serialization support for the `GlobalAlignmentKernelKMeans`, `TimeSeriesKMeans`, `KShape`, `KNeighborsTimeSeries` and `KNeighborsTimeSeriesClassifier`, `PiecewiseAggregateApproximation`, and `SymbolicAggregateApproximation` models, enabling models to be saved to various formats (HDF5, JSON, and pickle) and subsequently loaded. The commits also included tests to verify the correct serialization and deserialization of these models, ensuring reproducibility and usability of the trained models.
machine-learningpythontime-seriestimeseriestime-series-analysis
fastplotlib/fastplotlib
Apr 2022 - Jan 2023
Next-gen fast plotting library running on WGPU using the pygfx rendering engine
Contributions:21 releases, 464 reviews, 302 commits in 9 months
rendering-enginegpuscientific-visualizationvisualizationvulkan
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.