Research Scientist (Robotics & World Models)

Location: Remote - United States

Salary: $160,000–$185,000 per year, based on experience, skills, and qualifications.

Company

hirify.global is a global data engineering company providing data, evaluation frameworks, platforms, and human expertise for generative AI and other AI systems.

What you will do

• Define, structure, and experimentally validate data for robot foundation models, including vision-language-action models, world models, and manipulation and locomotion policies.
• Translate model requirements into data specifications covering modalities, action representations, sampling, annotation schemas, and evaluation criteria.
• Curate and weight datasets across robot embodiments, action spaces, sensors, real-world capture, simulation, and synthetic-data pipelines.
• Build benchmarks for coverage, reliability, discriminative power, world-model rollout quality, and sim-to-real transfer.
• Fine-tune and evaluate foundation models using data-quality ablations and scaling studies, and design adversarial and long-horizon evaluations.
• Collaborate with capture, annotation, and synthetic-data teams while publishing benchmarks, methodologies, and research papers.

Requirements

• Approximately 4+ years of hands-on industry experience in robot learning or robotics machine learning; a current PhD research agenda may offset the lower end.
• Bachelor's degree in computer science, electrical engineering, robotics, or a related technical field; an MS or PhD is preferred.
• Experience training and evaluating manipulation or locomotion policies with imitation learning or reinforcement learning, with strong PyTorch fundamentals.
• Experience curating and weighting robot datasets across embodiments and sensor setups, using formats such as LeRobot, RLDS, and Open X-Embodiment.
• Hands-on experience with simulation and synthetic data, including Isaac Sim, Isaac Lab, Omniverse, MuJoCo, domain randomization, system identification, and sim-to-real transfer.
• Experience with teleoperation or egocentric data collection, VLM adaptation, VLA fine-tuning, Hugging Face Transformers, PEFT, and recognized research publications or open-source contributions.

Nice to have

• Interest or hands-on experience in responsible-AI evaluation and red-teaming for embodied systems, including safety and robustness testing.

Culture & Benefits

• Direct collaboration with customers and frontier research labs developing robot foundation models.
• Practical, research-driven work focused on rigorous, reproducible experiments and documentation.
• Opportunity to publish benchmarks, methodologies, and papers that advance robotics and AI.