Modeling And Simulation Engineer Jobs in Berkeley, CA

Introducing Moonlake, AI for creating world simulations.

Moonlake is building the frontier of AI-powered world simulation.

We create systems that generate, simulate, and reason over rich 3D environments for robotics, embodied AI, and interactive applications. Our platform enables the creation of digital worlds, synthetic environments, and scalable simulation infrastructure used to train the next generation of intelligent systems.

Our work sits at the intersection of:

• Robotics
• Physical AI
• World Models
• Simulation Infrastructure
• Synthetic Data Generation
• Embodied Intelligence

Moonlake has raised $28M in seed funding from NVIDIA Ventures, Threshold Ventures, AIX Ventures, and notable angels including Naval Ravikant and Jeff Dean.

Our mission is to build the foundational infrastructure that enables robots to learn, reason, and operate effectively in the physical world.

We are looking for a Member of Technical Staff – Robotics to help build the bridge between simulation, world models, and real‑world robotic systems.

This role spans the full robotics stack—from evaluating foundation models and policies in simulation, to training world models, to deploying and operating physical robots. You will work closely with researchers and engineers developing next‑generation simulation environments and AI systems, while ensuring those capabilities transfer successfully into real‑world robotic platforms.

This is a highly hands‑on role combining robotics engineering, machine learning, simulation, and hardware deployment.

• Benchmark and evaluate robot foundation models in simulated environments
• Design evaluation frameworks for robotic reasoning, planning, manipulation, and navigation
• Measure generalization, robustness, and task performance across diverse scenarios
• Build infrastructure for large‑scale simulation‑based testing and validation

• Develop and train world models that enable robots to understand and predict environment dynamics
• Build systems that learn from multimodal robot data including vision, depth, state, and actions
• Improve environment understanding, forecasting, and decision-making capabilities
• Work closely with simulation and AI teams to advance robotic world modeling systems

• Collect and curate real‑world robotics datasets
• Train and fine‑tune models using both simulated and physical robot data
• Improve sim‑to‑real transfer for robotic policies and world models
• Develop workflows connecting simulation, training infrastructure, and deployed robotic systems

• Set up, integrate, and maintain robotic hardware platforms
• Bring learned policies and world models onto real robotic systems
• Debug hardware, software, sensing, and control issues
• Develop deployment pipelines for testing, validation, and continuous improvement
• Work directly with robotic manipulators, mobile robots, sensors, and compute systems

• Policy evaluation
• Model benchmarking
• Simulation‑based testing
• Generalization analysis
• Performance measurement

• Environment modeling
• Predictive systems
• Representation learning
• Multimodal learning
• Model‑based reasoning

• Robotics simulators
• Digital twins
• Synthetic environments
• Sim‑to‑real transfer
• Evaluation infrastructure

• Robot setup and integration
• Sensors and perception systems
• Robot control
• Hardware debugging
• Deployment workflows

• Strong background in robotics, embodied AI, machine learning, or related fields
• Experience working with physical robotic systems
• Experience with robotic simulation platforms such as Isaac Sim, MuJoCo, Habitat, Gazebo, or similar
• Familiarity with robot learning, foundation models, or world models
• Strong software engineering skills in Python and robotics tooling
• Experience deploying software onto real robotic hardware
• Ability to debug across hardware, software, and machine learning systems
• Comfort working in a fast‑moving research and engineering environment

Moonlake's vision extends beyond simulation. We believe the future of robotics will be powered by world models that can learn in simulation and transfer seamlessly to the physical world.

This role sits at the center of that mission. You will help evaluate robotic intelligence in simulation, train the models that power robotic understanding, and deploy those systems onto real robots operating in the physical world.

Your work will directly shape how future robotic systems learn, reason, and act.

We are committed to being an on‑site, in‑person team currently based in San Francisco.