Magic
Member of Technical Staff, Pre-training Systems
San Francisco · Staff+
Sponsorship not specified$225k-$550kDetected 144 days ago
Node.jsDistributed SystemsMachine LearningDeep LearningMLOpsResearchCommunication
About the role
- This role focuses on large-scale model training across massive GPU clusters.
- You will work at the boundary between deep learning and distributed systems, ensuring that training runs are performant, reliable, and reproducible under extreme scale.
- Scale distributed training across large GPU clusters (data, tensor, pipeline parallelism)
Responsibilities
- As a Software Engineer on the Pre-training Systems team, you will design and operate the distributed infrastructure that trains Magic's long-context models at scale.
- You will own the systems that make large-scale pre-training stable and fast.
- Collaborate with Kernels and Research to align model architecture with systems realities
Requirements
- Experience training large models in multi-node GPU environments
- Experience debugging cross-layer issues in production ML systems
- Strong ownership mindset and ability to operate critical infrastructure
- Track record of improving performance or reliability of large-scale systems
Compensation
- Equity is a significant part of total compensation, in addition to salary
- 401(k) plan with 6% salary matching
- A small, fast-paced, highly focused team
- Integrity. Words and actions should be aligned
- Hands-on. At Magic, everyone is building
- Teamwork. We move as one team, not N individuals
Benefits
- Generous health, dental and vision insurance for you and your dependents
- Unlimited paid time off
- Visa sponsorship and relocation stipend to bring you to SF, if possible
- We value quick learning and grit just as much as skill and experience.
Company info
- Magic's mission is to build safe AGI that accelerates humanity's progress on the world's most important problems.
Visa & Work Authorization
- Visa sponsorship and relocation stipend to bring you to SF, if possible
This listing is sourced directly from Magic's careers page and normalized into a canonical job model.