Sharad Vikram - Senior Staff Software Engineer at Google DeepMind

Sharad Vikram

Senior Staff Software Engineer at Google DeepMind

San Francisco, California, United States

Join Prog.AI to see contacts

Summary

🤩

Rockstar

🎓

Top School

Sharad Vikram is a Senior Staff Software Engineer and ML researcher at Google DeepMind with 14 years of experience building core ML infrastructure and algorithms, currently focused on JAX core, the Pallas kernel language, and Gemini LLM work. He has deep systems-and-algorithms expertise, contributing to high-impact open-source projects like JAX, Triton, XLA, and TensorFlow Probability—work that ranges from implementing jax.numpy functionality and SVD numerical fixes to enabling Python callbacks in XLA and extending Triton’s frontend. Comfortable at the intersection of research and engineering, he ships production-grade kernel and compiler improvements that accelerate GPU/TPU workloads. Based in San Francisco, he pairs a PhD-level ML background with hands-on backend systems design, and is known for finding numerical-stability and interoperability gaps that unlock real-world ML performance.

14 years of coding experience

9 years of employment as a software developer

BS Electrical Engineering and Computer Science, BS Electrical Engineering and Computer Science at University of California, Berkeley

University of California, San Diego

Tamil

Github Skills (22)

c-language10

python10

numpy10

triton10

compiler-design10

tensorflow10

compiler10

jax10

cprogramming-language10

unit-test10

linear-algebra10

gpu-programming9

front-end-development9

cuda9

apidoc8

Programming languages (8)

C++JavaScriptHaskellLuaVim scriptMLIRJupyter NotebookPython

Github contributions (5)

jax-ml/jax

Nov 2019 - Jan 2023

Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more

Role in this project:

Back-end Developer

Contributions:3 releases, 506 reviews, 164 commits in 3 years 2 months

Contributions summary:Sharad primarily contributed to implementing the `tile` functionality within the `jax.numpy` module, demonstrating involvement in core library development. This involved writing the necessary implementation, creating tests, and updating documentation to reflect the new feature. Furthermore, the user's work extended to optimizing linear algebra routines, including addressing numerical stability issues in the Singular Value Decomposition (SVD) algorithm, as evidenced by modifications and improvements to relevant tests.

pytorchpythonjitautomatic-differentiationgpu

tensorflow/probability

Apr 2019 - Jul 2022

Probabilistic reasoning and statistical analysis in TensorFlow

Role in this project:

Back-end Developer, ML Engineer

Contributions:250 commits, 5 comments in 3 years 3 months

Contributions summary:Sharad's commits primarily focused on enhancing the JAX NumPy implementation within the TensorFlow Probability library. They introduced and implemented the `lax numpy.tile` function, added documentation for the `tile` function, and addressed issues within the singular value decomposition (SVD) functionality. Furthermore, the user contributed to the improvement of machine learning-related aspects through code adjustments for complex numbers, along with general tests and custom gradients.

statisticspythonprobabilistic-reasoningdata-sciencedeep-learning

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.

Request Free Trial