RL Evaluation Engineer - Scalable Distributed Systems

SpaceX Super Intelligence; Palo Alto (CA); On-site; USD 180,000 - 440,000; Full time

Benefits offered by this job

Equity

Medical coverage

Vision coverage

Dental coverage

401k plan

Disability insurance

Life insurance

Various discounts

Job summary

SpaceXSI is seeking an engineer to help develop our evaluation infrastructure. The RL infrastructure team is building scalable systems to run evaluations at scale and collaborating with modeling and evaluation teams on new evals and improving signal quality.

You will identify bottlenecks across the stack—from inference to scheduling and orchestration—driving dynamic resource optimization and tooling to support the next generation of models.

Qualifications

• Experience building, debugging, and optimizing efficiency of large-scale distributed systems.
• Willingness to dive deep and solve hard problems at all levels of the stack.
• Experience in LLM inference and GPU/compute management is preferred.

Responsibilities

• Build highly reliable, efficient, and easy to use infrastructure that runs all our evaluations at scale
• Collaborate closely with modeling & evaluation teams on developing new evals, maintaining evaluation signal quality, and building internal tooling to support training our next-generation of models
• Identify and resolve performance bottlenecks in all layers of the eval infrastructure stack, including inference, capacity fleet management, asynchronous eval orchestration, and more

Skills

Distributed systems

LLM inference

GPU scheduling

Resource optimization

Model eval interfaces

Job description

SpaceXSI is seeking an engineer to help develop our evaluation infrastructure. The RL infrastructure team is building scalable systems to run evaluations at scale and collaborating with modeling and evaluation teams on new evals and improving signal quality.

You will identify bottlenecks across the stack—from inference to scheduling and orchestration—driving dynamic resource optimization and tooling to support the next generation of models.