Harsh Bafna
Senior Member Of Technical Staff at Unravel Data
Chhattisgarh, India
Summary
🤩
Rockstar🎓
Top SchoolHarsh Bafna is a Senior Member of Technical Staff with 6 years of experience building production-grade ML model serving and big data platforms, currently at Unravel Data. He is a top contributor to PyTorch's TorchServe project, driving backend refactors and protocol improvements that helped consolidate the framework around PyTorch. Harsh blends deep ML-serving expertise (TorchServe, Triton, TF Serving) with hands-on cloud and analytics architecture across AWS services, Spark, Hive, and end-to-end ETL pipelines. He also brings practical DevOps and container experience (Docker, EKS, CircleCI) and has implemented gRPC APIs for scalable systems. A prolific problem-solver, he has 1900+ Stack Overflow reputation and a track record of owning design-to-release cycles in open-source and enterprise settings. Based in Chhattisgarh, India, he pairs backend systems craftsmanship with a knack for translating research-grade ML tooling into robust production services.
6 years of coding experience
14 years of employment as a software developer
Bachelor of Engineering (B.E.) Information Technology, Bachelor of Engineering (B.E.) Information Technology at Bhilai Institute of Technology Durg
English, Hindi
Stackoverflow
2,154reputation
207kreached
73answers
5questions
amazon-s3
top-5%
amazon-web-services
top-5%
Github Skills (16)
pytorch10
back-end-development10
python10
apidoc9
amazon-s39
amazon-web-services9
refactoring9
api9
amazon-rds6
apache-spark6
aws-lambda6
amazon-athena6
amazon-iam6
pyspark6
aws-glue6
Programming languages (4)
Github contributions (5)
pytorch/serve
Nov 2019 - Dec 2020
Serve, optimize and scale PyTorch models in production
Role in this project:
Backend Developer
Backend DeveloperContributions:99 reviews, 762 commits, 84 PRs in 1 year 1 month
Contributions summary:Harsh's commits focused on refactoring and improving the backend code of the PyTorch Serve project, specifically in the model service worker component. They removed references to the MXNet framework, indicating a shift or consolidation towards PyTorch. Further commits reflect the modification and update of messages exchanged between frontend and backend components to enhance the overall performance.
cpupytorchpytorch-modelsservingin-production
harshbafna/multi-model-server
Nov 2019 - Feb 2020
Model Server for Apache MXNet is a tool for serving neural net models for inference
Contributions:2 PRs, 1 push, 1 branch in 3 months
mxnetservingdeep-learninginferencemachine-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.