Kam Kasravi
San Jose, California, United States
Summary
🤩
Rockstar🎓
Top SchoolKam Kasravi is a Deep Learning Software Engineer based in San Jose with 13 years of experience building and optimizing ML systems for production at scale. At Intel he focused on distributed training and inference across CPU/GPU clusters, contributed to an XAI toolkit and created operator bundles for oneAPI Jupyter notebooks, blending model engineering with platform automation. A longtime Kubeflow contributor and member of its Technical Advisory Committee, he helped shape multi-tenancy, kustomize overlays and CI/CD pipelines that harden ML deployments. His open-source work spans containerization and deployment automation—fixing Dockerfiles, MPI launches, and test harnesses in high-profile repos like Intel’s AI Reference Models and Kubeflow. Comfortable across TensorFlow, PyTorch and Kubernetes, he brings rare depth in both low-level profiling/debugging and pragmatic DevOps to bridge research models into reliable production services.
13 years of coding experience
26 years of employment as a software developer
Bachelor's degree, Applied Mathematics, Bachelor's degree, Applied Mathematics at University of California, Los Angeles
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (24)
kubernetes10
docker10
testing10
kustomize10
bash10
dockers10
cicd10
scala10
automation10
go10
kubernetes-pods10
stream-processing10
automations10
devops10
streaming10
Programming languages (14)
Github contributions (5)
gearpump/gearpump
Aug 2014 - Mar 2017
Lightweight real-time big data streaming engine over Akka
Role in this project:
Back-end Developer
Back-end DeveloperContributions:213 commits, 227 PRs, 63 pushes in 2 years 7 months
Contributions summary:Kam made several commits involving changes to the `TaskActor.scala` and `Graph.scala` files, indicating work on the core streaming engine logic. Their commits involved merging branches and addressing conflicts within the `gearpump/gearpump` repository. The code changes appear to be related to task management, partitioning, and message handling within the distributed streaming engine.
stream-processingbigdatastreaming-datastreamingbig-data
kubeflow/kubeflow
Mar 2018 - Oct 2019
Machine Learning Toolkit for Kubernetes
Role in this project:Backend & DevOps Engineer
Backend & DevOps EngineerContributions:89 commits, 141 PRs, 1183 comments in 1 year 7 months
Contributions summary:Kam primarily focused on cleaning up and refactoring the Libsonnet code within the Kubeflow core. They made changes to the JupyterHub configuration and added unit tests for it. The commits involved modifications to networking and access control including TFJob UI integration and Istio related deployments. Also added the ability to generate names to skip kubectl create.
pythondata-sciencenotebookmachine-learningmlops
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.