Ehsan Kermani
AI Product Manager at Modular
Vancouver, British Columbia, Canada
Summary
🤩
Rockstar🎓
Top SchoolEhsan Kermani is an AI product manager and machine learning systems engineer with 11 years of experience building ML infrastructure, MLOps, and production-grade deep learning applications across cloud, edge, and embedded environments. He combines hands-on expertise in PyTorch, TensorFlow, MXNet and TVM with systems programming in Rust and distributed optimization using Scala and Apache Spark, having authored the Spark-LP package. Ehsan has driven anomaly detection and RAG/semantic search products at scale, contributed notable fixes and operators to high-profile open-source projects like Apache TVM and TensorFlow Rust, and helped optimize large models at OctoML. His background includes applied research at AWS and Huawei—where he holds a granted patent for a cloud-edge computer vision framework—and leadership in bringing MLOps best practices to production teams. A longtime open-source advocate and former Vancouver Rust meetup co-organizer, he blends deep research instincts with pragmatic product thinking. He is currently shaping AI infrastructure and developer experience at Modular, focused on MAX and Mojo.
11 years of coding experience
11 years of employment as a software developer
Master's Degree Computer Science, Master's Degree Computer Science at The University of British ColumbiaBachelor's Degree Mathematics, Bachelor's Degree Mathematics at Sharif University of Technology
English, persian (native)
Stackoverflow
942reputation
40kreached
3answers
10questions
Github Skills (21)
tvm10
mxnet10
machine-learning10
onnx10
deeplearning-ai10
compiler-compiler10
deep-learning10
tensorflow10
compiler10
rust10
refactoring10
python9
relayer9
relay9
save6
Programming languages (13)
Github contributions (5)
apache/tvm
May 2018 - Oct 2022
Open deep learning compiler stack for cpu, gpu and specialized accelerators
Role in this project:
ML Engineer
ML EngineerContributions:10 reviews, 13 commits, 12 PRs in 4 years 5 months
Contributions summary:Ehsan contributed to the `apache/tvm` repository by exposing the clip operator to the MXNet frontend, enabling the integration of clip functionality within the TVM framework for MXNet models. They added a tutorial on the TOPI, which provides generic operations and schedules within TVM. They also added the GridSample operator to the ONNX frontend. They were also fixing the rust resnet example to correctly use the example in the project.
metalvulkancompilertensoropencl
tensorflow/rust
Apr 2019 - May 2019
Rust language bindings for TensorFlow
Role in this project:Back-end Developer
Back-end DeveloperContributions:8 commits, 3 PRs, 11 comments in 12 days
Contributions summary:Ehsan contributed to the `tensorflow/rust` repository, which provides Rust bindings for TensorFlow. Their work primarily involved code formatting and refactoring, improving code readability and consistency. Key changes include reformatting the `src/lib.rs` file and the `src/graph.rs` file, ensuring the code adheres to established style guidelines. Additionally, the user added support for the Half TensorType.
rustmachine-learningtensorflowrust-languagerust-lang
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.