Medior RAI AI Reinforcement Learning Engineer

About the role

We are looking for a Medior RAI AI Reinforcement Learning Engineer to join our on‑site engineering team in Prague. You will design reinforcement‑learning and post‑training methods that make robot policies robust and reliable, contributing to the core intelligence that powers our humanoid robots.

Key responsibilities

• Train and evaluate models on REVEL's human‑demonstration data.
• Build scalable, reproducible training and evaluation pipelines.
• Collaborate across the RAI, hardware and data teams.
• Push research results into production on the robot.

Required profile

• 3+ years of relevant professional experience.
• Track record of building and training RL policies such as PPO, SAC, or offline RL, with deep understanding of their inner workings.
• Experience applying RL to robotics or physical‑system control, including deploying and validating policies on real hardware.
• Strong understanding of sim‑to‑real transfer, including domain randomization, system identification, latency, actuator dynamics, and common failure modes.
• Comfortable using physics simulators such as Isaac Lab, MuJoCo, or equivalent to train and stress‑test policies.

Required skills

• Reinforcement Learning (PPO, SAC, offline RL).
• Robotics control and physical‑system dynamics.
• Sim‑to‑real transfer techniques.
• Physics simulators (Isaac Lab, MuJoCo).

What we offer

• Unlimited Paid Time Off with real flexibility.
• On‑site work in our Prague lab, with visa sponsorship and relocation support.
• Complimentary lunch daily, plus coffee, snacks and drinks.
• Top‑spec GPU workstation, cluster access, and hands‑on time with robots.
• Conference budget for select roles and opportunities to publish or contribute to open‑source where IP allows.