Mid-Level Research Engineer, Benchmarks

About the Role

Join a small, technical team at an early-stage AI evaluation infrastructure startup building benchmarks for frontier AI agents. You will own the design and implementation of rigorous evaluations for realistic, domain-specific workflows, helping technical teams understand how agent performance translates to practical settings.

What You'll Do

• Design, implement, and maintain benchmarks for evaluating frontier AI agents on domain-specific tasks.

• Collaborate with subject-matter experts to turn real workflows into realistic benchmark tasks and evaluation criteria.

• Build reliable infrastructure to run models and agents against evaluation tasks at scale.

• Develop metrics and analyses to assess benchmark difficulty, reliability, and failure modes.

• Validate how benchmark results relate to real-world evaluations and technical user needs.

• Write clear technical documentation and reports that communicate findings to research and engineering audiences.

What We're Looking For

• Two or more years of experience in research engineering or machine learning engineering, with a focus on AI benchmarks, evaluation infrastructure, or agent environments.

• Experience designing, implementing, and operating benchmarks or evaluation environments for AI agents or large language models.

• Proficiency in Python, Docker, and Linux environments.

• Experience building infrastructure to run AI models or agents reliably, and collaborating with subject-matter experts to define tasks and criteria.

• Experience analyzing workflows across technical or business domains, and developing metrics or validation studies for benchmark quality.

• Strong technical writing, attention to detail, and ability to work independently in unstructured environments.

• Experience with reinforcement learning pipelines, data generation, published work, or public benchmarks is beneficial.

Compensation & Benefits

Annual salary range: $100,000 to $170,000 USD. Visa sponsorship is available.

Location

On-site in Singapore, Singapore.