Member of Technical Staff

About the role

This role advances post-training from data curation to large-scale optimization to drive measurable capability gains in model performance. The hire will own the design and execution of data and training pipelines that convert raw information into actionable intelligence for broader use cases. Close collaboration with pre-training and post-training teams is essential to deliver step-function improvements in how large models reason and follow instructions. Contributions to reinforcement learning research will deepen the understanding of model behavior and enable more robust instruction following. The position demands ownership of ambitious research or engineering agendas that produce tangible, measurable improvements in model capabilities. You will push post-training frontiers by managing complex data curation strategies and optimizing large-scale training processes. Success requires fluency in navigating intricate ML codebases and distributed systems to ensure reliable execution at scale. You will translate cross-functional requirements into clear technical outcomes that align with strategic product and research goals.

Key facts

What you'll do

Design and implement data curation strategies that transform raw information into high-quality training signals for post-training workflows.

Develop and maintain large-scale optimization pipelines that enable measurable progress in model alignment and general agent capabilities.

Conduct research and engineering initiatives that push the frontiers of post-training, focusing on reinforcement learning and instruction-following behaviors.

Analyze complex model performance metrics and translate findings into concrete improvements for data, reward models, and training objectives.

Own ambitious research or engineering agendas that deliver tangible, measurable improvements across the model lifecycle.

Collaborate fluidly across research and infrastructure boundaries to ensure cohesive progress on demanding technical objectives.

Partner with pre-training and post-training teams to integrate insights and align on shared standards for model performance and reliability.

Apply strong engineering skills to dive into intricate ML codebases and distributed systems, ensuring stable and scalable execution.

Demonstrate clear communication and cross-functional collaboration skills when working with business teams and stakeholders.

Thrive in a fast-paced, high-agency startup environment by maintaining a bias toward action and clarity of execution.

Contribute to reinforcement learning research that deepens understanding of how large models reason, learn from feedback, and improve over time.

Support the development of open models that can be customized and built upon for societal benefit, aligning with open intelligence principles.

Leverage data pipelines, reinforcement learning frameworks, and inference optimization tools to drive impactful results.

Participate in typical data interview processes, including SQL or coding exercises, statistics questions, and case studies that assess analytical thinking.

Requirements

Deep understanding of machine learning fundamentals and practical experience with large-scale LLM training is required to navigate complex workflows and technical constraints.

Strong engineering skills enable diving into intricate ML codebases and distributed systems for reliable execution in demanding production environments.

Demonstrated ability to improve model behavior through data, reward modeling, or reinforcement learning techniques with tangible, measurable outcomes.

Capacity to own ambitious research or engineering agendas that produce measurable model improvements and advance post-training objectives.

Thriving in a fast-paced, high-agency startup environment demands a bias toward action, clarity of execution, and consistent cross-team coordination.

Working fluidly across research and infrastructure boundaries ensures alignment on goals, timelines, and quality standards for demanding projects.

Communication capabilities support effective collaboration with cross-functional stakeholders, including data scientists, engineers, and business teams.

A degree is required as part of the formal qualifications for this role within the research department and on-site work environment.

Nice to have

Team building activities include regular off-sites, happy hours, and celebrations to strengthen collaboration and team cohesion.

Practical notes

The role is based in San Francisco, California, within an on-site work environment.

Employment is full-time under an exempt research and technical staff arrangement.

Export Administration Regulations may apply, potentially requiring government authorization for offer confirmation and for specific expectations and processes.

Typical interview steps

Data interviews commonly include a SQL or coding exercise, a statistics question, and a case study. Candidates may be asked to design a metric, interpret an experiment, or build a small model. Some companies give a take-home analysis. Expect questions about past projects and the business impact of your work. Interviewers often evaluate how you communicate uncertainty and business impact, not only the math. Bringing a clean write-up of a past analysis to the interview is well received.