Kumar Agrawal
Member Of Technical Staff at University of California, Berkeley
San Francisco, California, United States
Summary
👤
Senior🎓
Top SchoolKumar Agrawal is a Member of Technical Staff and PhD candidate in Computer Science at UC Berkeley with 12 years of experience building algorithms and systems for human-centric machine learning, especially in vision, language, and long-context multimodal representation learning. He combines rigorous research at institutions like Google Brain, UCSF, and Mila with practical engineering—contributing QA and test automation to flagship projects such as the Julia language and improving core numerical and symbolic libraries like Theano and SymPy. Based in San Francisco, he focuses on scalable training and efficient inference, with applied work toward personalized cancer care and robotics-driven program synthesis. Comfortable across CS/systems, statistics/ML, and robotics, he brings a rare mix of formal mathematical training and production-quality software craftsmanship, demonstrated by deep test-suite enhancements and algorithmic fixes in widely used open-source projects.
12 years of coding experience
Doctor of Philosophy - PhD Computer Science, Doctor of Philosophy - PhD Computer Science at University of California, BerkeleyBS/MS Mathematics and Computing, BS/MS Mathematics and Computing at Indian Institute of Technology, Kharagpur
English, Hindi, German, French, Japanese
Github Skills (24)
computer-algebra10
programming-language10
python10
operation10
sympy10
testing10
tensorrt10
mathematics10
math10
computer-algebra-system10
tensorflow10
tensor10
unit-test10
theano10
julia10
Programming languages (9)
Github contributions (5)
Theano/Theano
Jul 2016 - Aug 2016
Theano was a Python library that allows you to define, optimize, and evaluate mathematical expressions involving multi-dimensional arrays efficiently. It is being continued as PyTensor: www.github.com/pymc-devs/pytensor
Role in this project:
Back-end Developer
Back-end DeveloperContributions:6 commits, 1 PR, 3 comments in 1 month
Contributions summary:Kumar primarily focused on improving the Theano library's functionality and stability. Their contributions include implementing a corrected two-pass algorithm for variance calculation and adding support for degrees of freedom (ddof). They also addressed pep8 errors and updated the documentation, while also adding and updating tests to ensure the correctness of the implemented features.
python-librarymathmulti-dimensionalpythonevaluate
sympy/sympy
Aug 2015 - Aug 2015
A computer algebra system written in pure Python
Role in this project:Back-end Developer
Back-end DeveloperContributions:6 commits, 3 PRs, 5 comments in 12 days
Contributions summary:Kumar primarily contributed to the `sympy/sympy` repository by modifying the `solveset` module and related testing files. Their work involved converting failing tests to pass and refining the behavior of `ConditionSet` objects, specifically related to handling equations and inequalities within the solving framework. The changes also incorporated support for inequalities and involved refactoring code to improve the overall functionality of the solvers.
mathpythonsciencecomputer-algebra-systemalgebra
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.