Madhusudan Jarhad
Architect at Persistent Systems
Pune District, Maharashtra, India
Summary
🤩
Rockstar🎓
Top SchoolMadhusudan Jarhad is an architect and engineering leader with over 6 years of focused professional experience and a long track record in software delivery across major Indian tech services firms. He currently drives architecture at GlobalLogic after leading projects at Persistent Systems, bringing practical expertise in backend systems, tooling integration, and production deployments. An active contributor to the PyTorch Serve project, he has improved CLI/versioning, Homebrew packaging, and benchmarking workflows—showing a flair for improving developer experience and deployability of ML services. Comfortable spanning hands-on coding and technical leadership, he has a background in computer science from the University of Pune and deep operational knowledge from large enterprise engagements. Colleagues describe him as a pragmatic problem-solver who notices small UX and integration gaps that make a big difference in production. He combines steady delivery in legacy environments with modern open-source contributions that bridge ML tooling and production systems.
6 years of coding experience
8 years of employment as a software developer
science, First, science, First at Vidyadham PrashalaSavitribai Phule Pune UniversityJawahar Navodaya Vidyalaya (JNV)
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (16)
command-line-interface10
pytorch10
serve10
python10
cli10
command-line10
argparse9
homebrew8
benchmark7
benchmarking7
docker6
dockers6
kubernetes-pods5
kubernetes5
machine-learning4
Programming languages (2)
Github contributions (5)
pytorch/serve
Mar 2020 - Jul 2020
Serve, optimize and scale PyTorch models in production
Role in this project:
Back-end Developer
Back-end DeveloperContributions:11 commits, 3 PRs, 11 pushes in 3 months
Contributions summary:Madhusudan primarily contributed to the TorchServe project by adding and modifying functionality related to the application's versioning and command-line interface. This includes adding version flags and support for them, as well as implementing changes for Homebrew installation and the use of JMeter for benchmarking. They also made adjustments to error messages and worker scaling responses and fixed the response status messages. Their contributions indicate a focus on improving user experience and integration with other tools.
cpupytorchpytorch-modelsservingin-production
mjpsl/tvm
Dec 2019 - Feb 2020
Open deep learning compiler stack for cpu, gpu and specialized accelerators
Contributions:4 PRs, 65 pushes, 10 branches in 2 months
cpugpu-programminggpu-accelerationtvmdeep-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.