Robot Learning Engineer

At FireLoop, we're building the next generation of Physical AI infrastructure, helping robotics teams bridge the gap between simulation and real-world deployment. By combining synthetic data generation, robot learning, and advanced simulation technologies, we enable faster development of intelligent robotic systems. The robotics industry is experiencing unprecedented growth, driven by breakthroughs in foundation models, synthetic data, and embodied AI. As a Robot Learning Engineer at FireLoop, you'll work on cutting‑edge technologies that help robots learn, adapt, and operate in the real world—contributing directly to the future of intelligent automation. The Role You'll own the full loop from synthetic data to deployed policy: the Isaac Lab/Mimic/Cosmos pipelines that generate and augment training data, and the models—VLA fine‑tunes, custom vision encoders, diffusion/flow‑matching action heads—that train on it. What You’ll Do Design and maintain SDG pipelines in Isaac Lab and Isaac Mimic, including scene and asset setup, demonstration generation and multiplication, domain randomization, and sensor/observation configuration. Integrate Cosmos (or similar world‑foundation‑model tooling) to improve visual realism of synthetic data and reduce the sim‑to‑real gap. Build and tune co‑training pipelines that combine real demonstration data with synthetic data, owning data ratios, curation, and evaluation. Fine‑tune existing Vision‑Language‑Action (VLA) models (e.g., pi0) on custom tasks and robot embodiments. Design and train custom, task‑specific vision encoders and action heads, including diffusion policies and flow‑matching policies for visuomotor tasks. Perform sim‑to‑real validation by diagnosing transfer failures and iterating on both model architectures and data generation strategies. Build and maintain training infrastructure, including data loaders, logging systems, evaluation harnesses, and ablation tooling to accelerate experimentation. What We Need 3-4 years of hands‑on experience training deep learning models for computer vision, robotics, or related domains. Strong PyTorch expertise, including custom architecture implementation, training debugging, optimization, and loss‑function design. Experience with imitation learning and behavior cloning, ideally including diffusion or flow‑matching policies. Strong Python engineering skills with a focus on maintainable, testable, and production‑quality code. Experience with robotics simulators such as Isaac Sim/Lab, Isaac Gym, MuJoCo, PyBullet, or similar platforms. A strong interest in owning both the data pipeline and model development lifecycle. Ability to independently investigate problems, run experiments, and communicate findings effectively. Nice to Have Hands‑on experience with Isaac Lab, Isaac Mimic, or NVIDIA Omniverse. Experience using NVIDIA Cosmos or similar world foundation models for synthetic data generation and augmentation. Experience fine‑tuning Vision‑Language‑Action models such as pi0, OpenVLA, or similar architectures. Familiarity with USD scene composition, domain randomization frameworks, and GPU‑parallelized physics simulation. Research publications or open‑source contributions in robotics and robot learning (CoRL, RSS, ICRA, NeurIPS, LeRobot, Isaac Lab, etc.). How We’ll Evaluate You 1. AI Technical Screening The first round will be an AI‑based interview focused on robotics, machine learning, deep learning, reinforcement learning, imitation learning, computer vision, and related concepts. The AI interviewer will assess your technical knowledge and problem‑solving approach through a series of domain‑specific questions. 2. Practical Assessment A hands‑on exercise designed to evaluate your ability to work with robot learning workflows, model training, debugging, or simulation‑related tasks in a practical setting. 3. Technical Interview A deep technical discussion with the engineering team, covering your previous projects, architecture decisions, model training experience, robotics simulators, and real‑world problem‑solving approaches. 4. HR Discussion A final conversation to discuss your experience, career goals, cultural fit, and other role‑related details. #J-18808-Ljbffr 9series