Principal Architect - Reinforcement Learning (Physical AI)

Location: Full-time onsite at the Irving, Texas office five days a week; domestic relocation assistance is available. Visa sponsorship is available for eligible applicants.

Salary: $147,760–$240,110 per year.

Company

Build software, connectivity, autonomy, and AI solutions that connect machines, data, and people for safer, more productive, and sustainable construction operations.

What you will do

• Own the reference architecture for learning-based autonomy and define how tasks, observations, actions, rewards, curricula, policies, and models are structured.
• Architect reinforcement learning, imitation learning, and foundation-model post-training pipelines for perception, prediction, planning, and control policies.
• Design large-scale closed-loop training using simulation, digital twins, GPU-parallel environments, and neural or hybrid simulation.
• Define sim-to-real transfer, reward and metric design, policy evaluation, regression testing, and safety-case methodology for machine deployment.
• Set standards for reproducibility, experiment rigor, benchmark hygiene, and scalable training and evaluation in partnership with architecture and operations teams.
• Lead technology evaluation and deployment, develop application architecture deliverables, and collaborate with cross-functional teams to deliver business results.

Requirements

• Knowledge of requirements analysis tools, methods, and techniques, with the ability to elicit, analyze, and document functional and non-functional requirements.
• Knowledge of analytical techniques and tools, with the ability to identify root causes and create alternative solutions.
• Knowledge of platform architecture, processing mechanisms, application-system roadmaps, and supportive end-user interfaces.
• Knowledge of target architecture and the ability to align IT blueprints, roadmaps, architecture, and processes with business strategies.
• Understanding of effective communication concepts and the ability to accurately exchange ideas, information, and needs.
• Typically 8+ years in robot learning with reinforcement learning, or a PhD plus 5+ years, taking algorithms from development to deployment on real hardware.
• Experience with PPO, SAC, GRPO, model-based or world-model reinforcement learning, reward shaping, curriculum design, and compounding-error handling in closed-loop training.
• Experience with behavior cloning, offline RL including IQL and GAIL, distillation, diffusion policies, and vision-language-action or robot foundation models with RL-based post-training such as RLHF, DPO, or GRPO.
• Experience architecting GPU-parallel simulation frameworks using NVIDIA Isaac Lab, Isaac Sim, MuJoCo, or Brax with Ray, Kubernetes, or Slurm; experience with neural or hybrid simulation such as NVIDIA Cosmos is required for the stated specialization.
• Experience with domain randomization, system identification, actuator and latency modeling, real-to-sim reconstruction, and field-task measurement for sim-to-real transfer.
• Experience defining learned-policy metrics, statistically rigorous comparisons, continuous evaluation and regression, adversarial and long-tail testing, runtime safeguards, and safety-case methods including ISO 26262, SOTIF, or ISO 13849.
• Knowledge of MPC, impedance control, trajectory optimization, learned versus classical approaches, and real-time perception-to-action pipelines.
• Experience shaping demonstration and fleet-data collection, auto-labeling, and synthetic-data strategies for imitation and offline RL.
• Strong Python and PyTorch skills, working C++ for real-time integration, and familiarity with CUDA, experiment tracking, Bazel or monorepo tooling, and CI/CD.
• Experience turning research into repeatable engineering, setting reproducibility and benchmark standards, mentoring ML engineers and researchers, and shipping a learning-based capability.
• Ability to work onsite full-time at the Irving, Texas office five days per week.

Nice to have

• JAX experience.
• Hydraulic or heavy-machinery dynamics experience.
• Off-highway, construction, or mining autonomy experience, or robot-learning experience at organizations such as NVIDIA, Waabi, Wayve, Zoox, Figure, 1X, Agility, Boston Dynamics, or Tesla Optimus.
• Master’s or doctoral degree in Computer Science, Robotics, Machine Learning, Controls, or a related field.
• Research or publication record.

Culture & Benefits

• Work on AI systems for autonomous and AI-enabled construction equipment using data from more than 1.6 million connected assets.
• Medical, dental, and vision benefits, paid time off, holidays, and volunteer time subject to plan eligibility.
• 401(k), HSA, FSA, disability, life insurance, parental leave, and adoption benefits.
• Career development, tuition reimbursement, incentive bonus, employee discounts, and employee assistance programs.