Aakash Nain
New Delhi, Delhi, India
Summary
🤩
Rockstar🎓
Top SchoolAakash Nain is a Senior Machine Learning Engineer with a decade of experience building and shipping production-grade deep learning systems, currently at Emergence AI after senior roles at H2O.ai and Ola. He combines research-oriented rigor with practical engineering—contributing to flagship open-source projects such as Keras and TensorFlow Addons (implementing initializers, WGAN-GP training logic, focal loss, GeLU and Cohen’s Kappa) which highlights his influence on the broader ML ecosystem. Comfortable across vision, generative models, and explainability, he routinely implements custom training loops and metrics that bridge research and deployable code. Based in New Delhi, he brings a Pythonic, collaborative approach to ML infrastructure and model tooling, with a track record of turning novel algorithms into well-tested library components.
11 years of coding experience
8 years of employment as a software developer
Bachelor of Technology (B.Tech.), Computer Science and Engineering, Bachelor of Technology (B.Tech.), Computer Science and Engineering at Guru Gobind Singh Indraprastha University
Hindi, English
Github Skills (14)
neural-network10
keras10
computer-vision10
eval10
deeplearning-ai10
machine-learning10
deep-learning10
tensorflow10
loss10
artificial-neural-networks10
python10
generative-adversarial-network10
evaluation10
jax4
Programming languages (10)
Github contributions (5)
tensorflow/addons
Mar 2019 - Jul 2020
Useful extra functionality for TensorFlow 2.x maintained by SIG-addons
Role in this project:
ML Engineer
ML EngineerContributions:58 reviews, 13 commits, 63 PRs in 1 year 3 months
Contributions summary:Aakash implemented and tested a focal loss function and included it within the TensorFlow Addons library. They added a GeLU activation layer, including comprehensive tests, and made it compatible with various data types. The user also incorporated the Cohen's Kappa metric, contributing to the library's capabilities for model evaluation.
tensorflowmachine-learningdeep-learningneural-networktensorflow-addons
keras-team/keras-io
May 2020 - Dec 2022
Keras documentation, hosted live at keras.io
Role in this project:ML Engineer
ML EngineerContributions:28 reviews, 7 commits, 6 PRs in 2 years 7 months
Contributions summary:Aakash implemented a WGAN-GP model within the Keras framework, adding a generative adversarial network for image generation. The contributions involve defining the discriminator and generator models, and overriding the `train_step` function for custom training logic, including gradient penalty calculation. This user also implemented an Integrated Gradients example, an explainability technique.
keras
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.