Founding Research Engineer

Scion Staffing, Inc.; San Francisco (CA); On-site; USD 150,000 - 200,000; Full time

Benefits offered by this job

Founding Engineering Opportunity

Cutting-Edge AI Work

Leadership Growth

Job summary

Scion Technology is seeking a Founding Research Engineer to join an early-stage AI startup in San Francisco, onsite, driving RL research and post-training evaluations. You will work directly with frontier AI labs to create RL tasks, build benchmarks, and evaluate cutting-edge AI models, with substantial ownership of technical projects and the potential to lead as the company grows.

The role requires hands-on engineering, research-driven experimentation, and the ability to translate complex

Qualifications

• Strong foundational understanding of reinforcement learning, coding RL, model evaluations, benchmarks, evaluation harnesses, and AI post-training.
• Ability to combine research intuition with hands-on engineering to design, build, and evaluate technical solutions.
• Experience with RL tasks, model evaluations, post-training workflows, or related AI research through professional work, academic research, or independent projects.

Responsibilities

• Design, develop, and implement coding reinforcement learning evaluations, tasks, and task creation pipelines.
• Build and enhance RealSuite, evaluating frontier AI models against benchmarks and representative coding tasks.
• Collaborate directly with leading frontier AI labs to develop RL tasks tailored to their research and training needs.
• Develop evaluation harnesses, benchmarks, and testing methodologies to assess model performance.
• Work closely with customers to understand technical requirements, ensure quality, and deliver solutions that meet project specifications.
• Combine research-driven experimentation with hands-on engineering to improve evaluation and post-training workflows.
• Take ownership of technical projects from initial research and design through implementation and delivery.
• Contribute to the growth of the engineering organization, with the opportunity to build and lead a team as the company scales.

Skills

Reinforcement learning

Model evaluations

Evaluation harnesses

Post-training AI workflows

Ownership of projects

Tools

RealSuite

Job description

Scion Technology has been engaged to conduct a search for an experienced Founding Research Engineer for our client, an early-stage AI startup. This is a full time, direct hire opportunity, onsite in San Francisco, CA.

POSITION OVERVIEW

We are seeking a highly motivated Founding Research Engineer to join an early-stage AI startup in San Francisco, working at the forefront of reinforcement learning (RL), model evaluations, and AI post-training. In this role, you will work directly with leading frontier AI labs to develop coding RL tasks, build benchmarks, and evaluate cutting-edge AI models. As an early engineering hire, you'll have significant ownership over technical projects, collaborate closely with the founders, and play a key role in shaping the company's engineering capabilities.

PERKS

• Founding Engineering Opportunity: Join as an early engineering hire with substantial ownership and influence over technical direction.
• Cutting-Edge AI Work: Collaborate directly with frontier AI labs on reinforcement learning, model evaluations, and post-training research.
• Leadership Growth: Opportunity to build and lead an engineering team as the company scales.

RESPONSIBILITIES

• Design, develop, and implement coding reinforcement learning (RL) evaluations, tasks, and task creation pipelines.
• Build and enhance RealSuite, evaluating frontier AI models against benchmarks and representative coding tasks.
• Collaborate directly with leading frontier AI labs to develop RL tasks tailored to their research and training needs.
• Develop evaluation harnesses, benchmarks, and testing methodologies to assess model performance.
• Work closely with customers to understand technical requirements, ensure quality, and deliver solutions that meet project specifications.
• Combine research-driven experimentation with hands-on engineering to improve evaluation and post-training workflows.
• Take ownership of technical projects from initial research and design through implementation and delivery.
• Contribute to the growth of the engineering organization, with the opportunity to build and lead a team as the company scales.

QUALIFICATIONS

• Strong foundational understanding of reinforcement learning (RL), coding RL, model evaluations, benchmarks, evaluation harnesses, and AI post-training.
• Ability to combine research intuition with hands-on engineering to design, build, and evaluate technical solutions.
• Experience with RL tasks, model evaluations, post-training workflows, or related AI research through professional work, academic research, or independent projects.
• Demonstrated technical initiative through experience at an AI data company, research organization, or self-directed projects involving benchmarks, RL, or post-training experimentation.
• Strong problem-solving skills and the ability to independently navigate complex technical challenges.
• High level of ownership, intellectual curiosity, and motivation to continuously learn and develop new skills.
• Ability to thrive in a fast-paced, early-stage startup environment with evolving priorities and significant individual responsibility.
• High integrity, strong work ethic, and a genuine passion for advancing AI research and engineering.

COMPENSATION & BENEFITS

This direct hire position offers a salary range of $150-200k, commensurate with experience. They offer a robust benefits package that includes comprehensive medical, dental, and vision coverage, along with additional perks.