Mark Sandler
Research Scientist at Google
Seattle, Washington, United States
Summary
🤩
Rockstar🎓
Top SchoolMark Sandler is a Research Scientist based in Seattle with 11 years of professional experience applying ML systems engineering at scale, currently working at Google. He has a strong research-to-production track record contributing core model work to high-profile open-source projects like TensorFlow (notably MobileNet V2 and EdgeTPU integration) and improving JAX and Colab tooling. His background blends a PhD in Computer Science with applied mathematics training, enabling both theoretical rigor and pragmatic optimizations around memory, sharding, and deployment. Colleagues rely on him for bug-forging fixes, performance tuning, and making research code production-ready across GPU/TPU stacks. An often-overlooked strength is his cross-cutting impact on developer experience—tests, documentation, and tooling improvements that smooth adoption of advanced ML features.
11 years of coding experience
1 year of employment as a software developer
Saint Petersburg State UniversityBachelor's degree, Applied Mathematics, Bachelor's degree, Applied Mathematics at Saint-Petersburg State University Information Technologies, Mechanic and Optics (University ITMO)State University of Nizhni Novgorod named after N.I. Lobachevsky (UNN)PhD, Computer Science, PhD, Computer Science at Cornell University
Github Skills (16)
web-widgets10
machine-learning10
tensorflow10
mobilenet10
jax10
widgets10
python10
google-colaboratory10
model-optimization9
numpy9
javascript9
data-serialization8
computer-vision8
json8
serialization8
Programming languages (5)
Github contributions (5)
googlecolab/colabtools
Nov 2017 - Jul 2021
Python libraries for Google Colaboratory
Role in this project:
Back-end Developer
Back-end DeveloperContributions:12 commits, 2 comments in 3 years 7 months
Contributions summary:Mark primarily focused on modifying and improving the `googlecolab/colabtools` repository, a collection of Python libraries for Google Colaboratory. Their commits show a shift in internal function calls, bug fixes, and the refactoring of javascript components. Moreover, the user implemented improvements for output area redirection, tab bar creation, and widget integration. They also made adjustments to the javascript encoder to support inheritance.
pythongoogle-colaboratorymachine-learningcolaboratorypython-libraries
tensorflow/models
Mar 2018 - Sep 2021
Models and examples built with TensorFlow
Role in this project:ML Engineer
ML EngineerContributions:9 reviews, 8 commits, 20 PRs in 3 years 6 months
Contributions summary:Mark primarily contributed to the TensorFlow models repository by implementing and modifying features related to MobileNet architectures. This includes the initial check-in of MobileNet V2, allowing configuration of batch normalization, and overriding NASNet model hyperparameters. Their work also involved open-sourcing training scripts and integrating MobilenetEdgeTPU models. Further contributions included bug fixes and code improvements, demonstrating a focus on model development and optimization within the TensorFlow ecosystem.
deep-learningtensorflow
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.