Model Engineer - Frontier RL & Large-Scale ML (Onsite PA)

Speedrun Talent Network; Palo Alto (CA); On-site; USD 160,000 - 250,000; Full time

Benefits offered by this job

Equity

Medical, Vision, and Dental coverage

401(k) retirement plan

Short & long-term disability insurance

Life insurance

Employee discounts and perks

Job summary

SpaceXSI, based in Palo Alto, CA, is hiring a Model Engineer to drive reinforcement learning and mid-training research in a small-team setting. You will own ambiguous problems end-to-end, form hypotheses, design experiments, build training/eval/data, and push results into next-gen models.

You will design rigorous experiments, improve data quality, define effective RL data and rewards, and advance training science from architecture to post-training recipes for coding and computer-use agents.

Qualifications

• Experience writing production-quality code in Python, PyTorch, JAX, Rust, or C++.
• MS or PhD in CS, ML, math, physics or a related field; equivalent research experience.

Responsibilities

• Design and run rigorous experiments that turn research ideas into measurable gains in model capability.
• Improve data recipes through data quality, deduplication, and mixture experiments.
• Define what makes RL data and rewards good, including grader design and preventing reward hacking.
• Advance training science, from architecture and scaling experiments to post-training recipes for coding and computer-use agents.
• Build evals and targeted tasks that expose model weaknesses and guide where the team invests next.
• Partner with engineers to take promising results from small experiments to frontier-scale training runs.

Skills

Python

PyTorch

JAX

Rust

C++

Education

MS or PhD in CS/ML/Math/Physics or related field

Tools

PyTorch (tool)

JAX (tool)

Job description

SpaceXSI, based in Palo Alto, CA, is hiring a Model Engineer to drive reinforcement learning and mid-training research in a small-team setting. You will own ambiguous problems end-to-end, form hypotheses, design experiments, build training/eval/data, and push results into next-gen models.

You will design rigorous experiments, improve data quality, define effective RL data and rewards, and advance training science from architecture to post-training recipes for coding and computer-use agents.