AI Engineer, RL & Evals

About the company

Our client is a seed-stage AI company building the data and infrastructure layer that improves model quality in subjective, creative domains, starting with design. It has raised $18.5M and has revenue-generating partnerships with frontier AI labs and application-layer companies. Its products support post-training, reinforcement learning, context and verification.

The role

As AI Engineer, RL & Evals, you will build reinforcement learning environments, evaluation frameworks and production infrastructure for creative domains. You will combine backend or full-stack product engineering with applied ML and own systems from design through delivery.

What you'll do

• Build and scale RL environments and evaluation frameworks for creative, non-verifiable domains.
• Design tasks and grading rubrics that capture design quality and taste.
• Build agent harnesses, context layers and supporting product infrastructure.
• Collaborate with internal research teams and frontier labs to improve model output quality.
• Ship production code end to end and build reliable backend systems and data pipelines.

Requirements

What we're looking for

• 2-7 years of relevant AI or reinforcement learning engineering experience, with depth building RL environments, evaluation systems or ML post-training systems.
• Strong full-stack or backend engineering skills and demonstrated production shipping experience.
• Experience with Python, PyTorch, RL, ML evaluation frameworks and LLMs.
• Knowledge of distributed systems and data pipelines; relevant scraping, indexing, embedding or agent-harness work is valuable.
• A computer science degree and fluent English communication.
• Ability to work in the San Francisco office five days per week.
• Bonus points
• Experience at creative technology, generative AI, crawling, indexing or data companies.
• Building agentic frameworks or evaluations for non-verifiable domains.
• Experience connecting research work to practical products and infrastructure.