Bandish Shah
Engineering Manager at Databricks
San Francisco, California, United States
Summary
🤩
Rockstar🎓
Top SchoolBandish Shah is a hardware-focused engineering leader and Member of Technical Staff in San Francisco with a decade-plus career designing SoC/ASIC solutions, FPGA-based systems, and high-speed I/O for enterprise computing. He has held senior roles from Principal Hardware Engineer at Oracle to leadership positions at SambaNova, MosaicML, and Databricks, blending deep systems architecture with people and program management. More recently he has bridged ML engineering and documentation work—contributing practical tutorials and profiler improvements to well-known MosaicML projects—demonstrating a knack for making complex tooling accessible. Trained in electrical engineering and systems engineering (Boston University, WPI), he combines low-level silicon experience with cloud and ML stack fluency. Colleagues know him for pragmatic architecture trade-offs and for translating research-grade hardware concepts into production-ready platforms.
4 years of coding experience
14 years of employment as a software developer
Masters Systems Engineering, Masters Systems Engineering at Worcester Polytechnic InstituteBachelor of Engineering - BE Electrical and Electronics Engineering, Bachelor of Engineering - BE Electrical and Electronics Engineering at Boston University
Github Skills (14)
pytorch10
machine-learning10
jupyter-notebook10
deep-learning10
python10
profiler10
documentation10
ml10
streaming10
datasets10
sphinx9
neural-network9
markdown-it9
markdown9
Programming languages (1)
Github contributions (5)
mosaicml/composer
Feb 2022 - Jan 2023
Supercharge Your Model Training
Role in this project:
ML Engineer
ML EngineerContributions:14 releases, 273 reviews, 124 commits in 11 months
Contributions summary:Bandish's commits primarily focused on refactoring and improving the `composer/profiler` module. The changes included restructuring the initialization process, streamlining profiling parameters, and updating references to timestamps. These modifications involved fixing tests and callback registration issues, demonstrating a focus on improving the efficiency and usability of the training profiler. Furthermore, the user worked on documenting the Profiler and sub-modules.
pytorchml-systemsdeep-learningneural-networksmachine-learning
mosaicml/streaming
Aug 2022 - Dec 2022
A Data Streaming Library for Efficient Neural Network Training
Role in this project:Technical Writer & ML Engineer (with a documentation focus)
Technical Writer & ML Engineer (with a documentation focus)Contributions:44 reviews, 7 commits, 19 PRs in 3 months
Contributions summary:Bandish primarily contributed to the documentation and example tutorials for the `mosaicml/streaming` repository, a data streaming library for efficient neural network training. They created and improved the documentation site, including the addition of a welcome page, quick start guide, and user guide. Their contributions involved creating and integrating examples such as CIFAR10 and FaceSynthetics tutorials in Jupyter notebooks, demonstrating the practical application of the library for ML tasks.
pytorchstreaming-datadeep-learningdatasetstreaming
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.