xAI The ML Infrastructure Engineer designs and scales GPU compute infrastructure and ML training frameworks to support rapid ML experimentation and productionization. This role collaborates with ML teams and mentors junior engineers.
Responsibilities
As an ML Infrastructure Engineer, you will play a pivotal role in building and optimizing the reliable, high-performance ML platform that powers recommendations on X. We're looking for exceptional engineers who are passionate about our mission and have a strong desire to make a meaningful impact.
Designing, building, and scaling GPU compute infrastructure, training frameworks, and experimentation tools to enable rapid iteration on ML hypotheses
Developing data pipelines and integrating large-scale data, training, and inference systems
Collaborating with ML teams to productionize models and ensure seamless integration across the stack
Ensuring scalability, reliability, and efficiency of large-scale machine learning systems
Working across the full stack to solve complex problems independently
Mentoring junior engineers and contributing to the growth of the team
Qualification
Bachelor2+ years experience with ML platformsExperience with job schedulers (e
Required
Bachelor, Master, Post-graduate or PhD in computer science, machine learning, or other quantitative discipline; or equivalent work experience
2+ years of industry experience working with high traffic or large-scale production environments, distributed systems, GPU infrastructure, and/or deep learning applications
2+ years experience with ML platforms, training infrastructure, or close collaboration with modeling engineers and data scientists
Strong proficiency with Python and experience with compiled languages such as C++ or Rust
Preferred
Deep familiarity with modern ML frameworks such as JAX or PyTorch
Low-level understanding of compute systems, including distributed storage, NVIDIA drivers, CUDA toolkits, and networking
Comfortable with Linux systems and orchestration tools
Experience with job schedulers (e.g., Slurm), configuration management (Puppet/Ansible), or related infrastructure tooling