Design, develop, and maintain the Python framework that enables algorithm developers across Mobileye to train, validate, quantize, and deploy deep learning models
locally, on-prem, and across cloud providers
through a single unified interface
Build high-performance data streaming libraries that feed large-scale distributed training pipelines in Rust with Python interfaces
Set the standard for reliable, reproducible research at scale
experiment tracking, configuration management, checkpoint handling, and multi-node training
Work directly alongside algorithm researchers to understand friction, propose solutions, and ship them
without layers of process in between
Contribute to open source when the right fix/feature belongs upstream
Requirements
A value-first mindset focused on shipping early and often
B.Sc. in Computer Science, Software Engineering, or equivalent hands-on experience
Strong software engineer skills in Python
tested, production-grade code that other engineers can build on
Familiarity with deep learning frameworks (ideally Pytorch) and distributed training workflows
Experience with containerization and CI/CD pipelines
Contributions to open source projects
Familiarity with Linux internals
networking, file systems, process management
Experience in Rust/C/Cuda
Experience with cloud infrastructure (AWS or similar) and distributed storage
Exposure to infrastructure-as-code or Kubernetes-based deployments
Tech Stack
AWS
Cloud
Kubernetes
Linux
Node.js
Open Source
Python
PyTorch
Rust
Benefits
Direct impact
the tools and infrastructure you build are used by every algorithm team at Mobileye, daily
A chance to work in one of Israel's largest AI factories
A small, highly technical team where everyone contributes across the stack and there are no ownership silos
Technically deep problems at the intersection of ML research, systems performance, and developer experience
The chance to work on problems where you are the last line