experiences

research, development, positions of responsibility, impact, ...

Below are my “formal” work experiences. I also contribute to several Open-Source Scientific Software, and I consider that as an “experience” too (see /opensource). I have also mentored several students and researchers who are new at writing code for research/scientific software (see /supervise).

TL;DR

Research Software Engineer intern

CERN, Princeton University
Starting January | Geneva, Switzerland
McMaster University
Prof. Jacques Carette
June - August 2023 | Hamilton, Ontario, Canada
  • Worked Type Theory, Functional Programming, and Computational Logic to formalize Mathematics and Computing in Agda’s (a dependently typed functional programming language and an interactive theorem prover) standard library using Type Theory, Proof Theory, Functional Programming, and Computational Logic.
  • Refactored the standard library’s codebase for v2.0, simplified the dependency graph, fixed various bugs, and added new functions and mathematical proofs.

Research Fellow

IRIS-HEP, Princeton Research Computing, Princeton University
Dr. Henry Schreiner (Princeton University / CERN), Dr. Jim Pivarski (Princeton University)
June - September 2022 | Remote
  • Work:
    • Directly responsible for preparing vector, a Python library for constructing Just-In-Time compilable jagged Lorentz vectors, for v0.9.0, v0.10.0, v0.11.0 and v1.0.0 (first major release).
    • Worked closely with researchers from CERN, IRIS-HEP (Analysis Systems group), and Princeton (Princeton Research Computing group) to develop new API, fix bugs, and build infrastructures for the HEP ecosystem. <!– - Impact:
    • Vector has 50+ GitHub stars and 210,000+ installs.
    • The releases are currently being used by researchers at CERN, Princeton University, IRIS-HEP and other research institutes. –>
  • Positions of responsibility, volunteering, talks:
    • Joined vector’s GitHub repository and Conda Feedstock as a maintainer.
    • Presented my work and vector at 21st International Workshop on Advanced Computing and Analysis Techniques in Physics Research, 5th International Workshop on Python in High Energy Physics, and multiple IRIS-HEP, Princeton Research Computing meetups (see /talks).
    • Invited to join PyHEP 2023.dev (first PyHEP developers meetup) at Princeton, but won’t be able to make it 😢
    • I still maintain/contribute to vector!

Technical Writer and Open-Source Developer

FluxML, Julia Programming Language
Mr. Dhairya Gandhi (Julia Computing / JuliaHub)
May - October 2022 | Remote (Part-Time)
  • Work:
    • Developed the documentation + packaging infrastructure and fixed bugs in FluxML, an ML and DL ecosystem that provides lightweight abstractions on top of Julia’s native GPU and AD support.
    • Wrote original Machine Learning/Deep Learning tutorials, documentation and API references for FluxML’s ecosystem. <!– - Impact:
    • FluxML is Julia’s primary ML and DL ecosystem with hundreds of thousands of downloads.
    • Flux.jl alone has 4000+ GitHub stars and 110,000+ installs (Julia-n ecosystems are not concentrated in a single library - for instance, there is a separate package under FluxML just for one-hot encoding - OneHotArrays.jl).
    • The documentation, infrastructue, and bug fixes impacted ML and DL researchers all around the world, including institutions and companies (MIT, AMD, UCL, CMU, and so on). –>
  • Recognition, talks, and more:
    • Joined FluxML’s GitHub organisation.
    • Will be presenting my work at JuliaCon 2023 (see /talks).
    • My work was shared on FluxML’s and JuliaLang’s official Twitter and LinkedIn accounts.
    • I still contribute to FluxML!

Google Summer of Code Developer

PyBaMM (Python Battery Mathematical Modeling), NumFOCUS
Dr. Valentin Sulzer (Carnegie Mellon University), Dr. Ferran Brosa Planella (University of Warwick), Dr. Robert Timms (University of Oxford)
May - September 2021 | Remote
  • Work:
    • Built an novel Twitter Bot (BattBot) capable of automatically constructing Mathematical Simulations of Batteries.
    • Developed new API, fixed bugs, and created new documentation for the PyBaMM ecosystem.
    • Created multiple build and packaging infrastructures for the PyBaMM ecosystem and the upstream packages. <!– - Impact:
    • PyBaMM is a collaboration between multiple academic institutes with 550+ GitHub stars and 300,000+ installs.
    • My work impacted battery researchers worldwide as PyBaMM is the most adopted Python framework for Modeling of Batteries.
    • BattBot gained a lot of traction on Twitter and GitHub. –>
  • Positions of responsibility, volunteering, talks: