Anand Subramanian
Machine Learning Engineer at EARTHBRAIN
Chiyoda, Japan
Summary
🤩
Rockstar🎓
Top SchoolAnand Subramanian is a Machine Learning Engineer with 10 years of experience specializing in deep learning, Bayesian analysis, and embedding ML into edge devices. He has built production-ready systems for 3D human body reconstruction, point-cloud perception for heavy machinery, and anomaly detection pipelines optimized to run on NVIDIA Jetson using TensorRT. Anand contributes to open-source ML tooling—authoring PyTorch Variational Autoencoder implementations with testing suites—demonstrating both research depth and engineering rigor. He blends academic training in deep learning and robotics with hands-on DSP and embedded systems experience from roles at IIT Madras and industry projects. Based in Chiyoda, Japan, he co-founded a local AI meetup to teach and apply ML practically, signaling a commitment to community building and knowledge transfer. Colleagues rely on him to translate probabilistic ML ideas into performant, deployable solutions across cloud and constrained hardware.
10 years of coding experience
8 years of employment as a software developer
Bachelor of Technology (B.Tech.), Mechatronics, Robotics, and Automation Engineering, CPGA 8.9/10, Bachelor of Technology (B.Tech.), Mechatronics, Robotics, and Automation Engineering, CPGA 8.9/10 at SRM UniversityMaster's degree, Deep Learning, Robotics, Master's degree, Deep Learning, Robotics at Japan Advanced Institute of Science and TechnologySindhi Model Senior Secondary School
English, Tamil
Github Skills (8)
variational-autoencoder10
pytorch10
deep-learning10
architectures10
modeling10
implement10
architecture10
machine-learning8
Programming languages (5)
Github contributions (5)
AntixK/PyTorch-VAE
Jan 2020 - Dec 2021
A Collection of Variational Autoencoders (VAE) in PyTorch.
Role in this project:
ML Engineer
ML EngineerContributions:72 commits, 5 PRs, 67 pushes in 1 year 11 months
Contributions summary:Anand implemented various Variational Autoencoder (VAE) models in PyTorch, as evidenced by the code differences. They developed a base VAE class and implemented specific VAE architectures, including a Vanilla VAE and potentially a Gamma VAE, along with the necessary training loop and loss functions, thus demonstrating an understanding of VAE principles and PyTorch implementation. The user also created a testing suite to validate model functionality and outputs, showcasing commitment to model correctness.
autoencoderspaper-implementationsceleba-datasetvaepytorch-vae
AntixK/Curve-Studio
Apr 2019 - May 2019
Contributions:1 release, 38 commits, 2 PRs in 13 days
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.