Job description

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. 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. Compensation and benefits • $175,000-$275,000 annual base salary. • Competitive equity. Location and work model Full-time, on-site five days per week in Levi's Plaza, San Francisco, California. Visa transfers, including OPT and H1B transfers, are supported.

Requirements

Must-haves

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.
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.

Nice-to-haves

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.

Application prep

Use the role’s language and application details while tailoring your materials.

Keywords to mirror

aimlpython

Platform: workable
Effort: medium