Staff Research Engineer - Frontier AI & RL Evaluation

Crossing Hurdles; United States; Remote; USD 150,000 - 210,000; Full time

Job summary

Crossing Hurdles is seeking a senior AI researcher to investigate frontier AI capabilities and training methods. You will formulate research questions guiding product strategy and collaborate with cross-functional teams to translate findings into practical applications.

The role emphasizes designing experiments, developing evaluation frameworks, and advancing synthetic data generation, reinforcement learning, and model understanding in a fast-paced research environment.

Qualifications

• Advanced degree in AI/ML or related field.
• 7+ years of relevant research experience.
• Strong Python and ML experimentation skills.
• Experience with synthetic or agentic data generation or RL.

Responsibilities

• Investigate capabilities, limitations, and training methods of frontier AI systems.
• Formulate research questions informing products, platforms, and technical strategy.
• Explore synthetic/agentic data generation, RL, post-training, model understanding, benchmarks, AI evaluation.
• Develop research-grade datasets, experiments, prototypes, tooling, and evaluation frameworks.
• Train, test, and evaluate models using modern AI and ML tools.
• Analyze results and derive evidence-based conclusions.
• Establish rigorous practices for experimental design, data quality, reproducibility, and interpretation.
• Collaborate with Research, Engineering, Product, and Operations to translate findings into practical applications.

Skills

Python

Research engineering

Machine Learning

AI research

Experiment design

Education

Masters or PhD in AI/ML/CS

Job description

Crossing Hurdles is seeking a senior AI researcher to investigate frontier AI capabilities and training methods. You will formulate research questions guiding product strategy and collaborate with cross-functional teams to translate findings into practical applications.

The role emphasizes designing experiments, developing evaluation frameworks, and advancing synthetic data generation, reinforcement learning, and model understanding in a fast-paced research environment.