Aditya Kane
Deep Learning Library Performance Software Engineering Intern at NVIDIA
Atlanta Metropolitan Area United States
Summary
🤩
Rockstar🎓
Top SchoolAditya Kane is a deep learning systems engineer and first-year PhD student at Georgia Tech with six years of experience accelerating AI through high-performance computing and CUDA kernels. Currently interning on deep learning library performance at NVIDIA, he has built fast multidimensional sliding-window attention implementations and contributed practical tutorials and layers to major open-source projects like Flax, Keras, and KerasCV. His work spans core library fixes, new computer-vision layers (StochasticDepth, DropPath, Squeeze-and-Excite), and reproducible educational notebooks, reflecting a focus on robust, test-backed contributions. Passionate about open source, he combines research rigor with production-minded engineering to make GPU-heavy ML primitives both faster and easier to use.
6 years of coding experience
4 years of employment as a software developer
Doctor of Philosophy - PhD Computer Science, Doctor of Philosophy - PhD Computer Science at Georgia Institute of TechnologyBachelor of Engineering - BE Computer Engineering, Bachelor of Engineering - BE Computer Engineering at Pune Institute of Computer Technology, Dhankawadi, Pune 411043
Stackoverflow
391reputation
96kreached
9answers
0questions
google-colaboratory
top-5%
Github Skills (19)
python10
testing10
machine-learning10
flax10
keras10
deep-learning10
tensorflow10
neural-network10
computer-vision10
jax10
data-science9
google-colaboratory9
optim7
opt7
data-pipeline6
Programming languages (13)
Github contributions (5)
keras-team/keras-cv
Jun 2022 - Nov 2022
Industry-strength Computer Vision workflows with Keras
Role in this project:
ML Engineer
ML EngineerContributions:58 reviews, 17 commits, 20 PRs in 5 months
Contributions summary:Aditya primarily contributed to the development of the `keras-cv` library, specifically focusing on implementing and testing computer vision layers. Their contributions included the addition of the `StochasticDepth`, `DropPath` and `SqueezeAndExcite2D` layers, as well as the integration of the `RandomResizedCrop` and `RegNet` models. This work involved modifying existing files and creating new ones to incorporate these features, as well as writing tests to ensure functionality.
visiondeep-learningcomputer-visionindustrystrength
keras-team/keras
Sep 2021 - Dec 2022
Deep Learning for humans
Role in this project:ML Engineer
ML EngineerContributions:26 reviews, 26 commits, 12 PRs in 1 year 2 months
Contributions summary:Aditya's commits primarily focus on modifying and improving the Keras library, a deep learning framework. They made changes to convolutional layers (Conv1D, Conv2D, Conv3D) to address invalid output shape issues and also fixed build errors. The user also added and corrected tests for locally connected layers and applications, ensuring the library's robustness and functionality. These changes indicate a focus on enhancing the core components and testing of deep learning models within Keras.
pythondata-sciencedeep-learningneural-networksmachine-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.