Senior Machine Learning Engineer (RL/World Models)

Salary: $94,000 - 134,000 per year Requirements:

• We are looking for a Masters or PhD in Computer Science, Robotics, Machine Learning, or a related discipline, or equivalent hands-on experience.
• We value experience developing and deploying reinforcement learning methods on real-world systems.
• We require strong proficiency in Python and deep learning libraries such as PyTorch or TensorFlow.
• We look for experience working with simulation environments such as MuJoCo or Isaac Gym.
• We expect a solid grasp of probability, statistics, and optimization.
• We prefer candidates who have trained and deployed machine learning models in production environments.

• Build action-conditioned world models that forecast how the welding process changes with robot motion and process settings.
• Learn the relationships among inputs, system state, physical dynamics, and weld quality.
• Develop multimodal models using video, 3D scans, thermal readings, electrical signals, robot state, and process parameters.
• Investigate latent dynamics, video prediction, generative modeling, and spatiotemporal representations.
• Improve long-range rollout accuracy, physical realism, temporal stability, and computational efficiency.
• Measure model uncertainty and determine when predictions are not trustworthy.
• Check learned predictions against real-world welding data.
• Integrate the model into reinforcement learning, planning, process optimization, evaluation, and synthetic data pipelines.
• Protect downstream optimization systems from taking advantage of model errors.
• Turn promising research into scalable training and inference systems.
• Develop reinforcement learning methods to improve welding decisions and process outcomes.
• Define state, observation, action, and reward structures around measurable manufacturing goals.
• Train and assess policies using learned world models, traditional simulation, offline datasets, and controlled real-world experiments.
• Build offline, model-based, or constrained RL techniques suited to limited and costly physical interaction.
• Optimize across tradeoffs such as weld quality, cycle time, reliability, energy consumption, and equipment constraints.
• Design approaches that handle uncertainty, distribution shift, delayed outcomes, and sparse or imperfect reward signals.
• Diagnose reward hacking, unsafe behavior, policy instability, and model misuse.
• Establish dependable offline and real-world policy evaluation methods.
• Collaborate with controls, welding, robotics, world-model, data, and ML infrastructure engineers.
• Convert research prototypes into reliable training, evaluation, and deployment systems.

• 3D
• AI
• Machine Learning
• PyTorch
• Python
• Robotics
• TensorFlow

More:

We are Path Robotics, and we are building the future of embodied intelligence with AI-driven robotic systems that can adapt, learn, and operate in the real world. Our work addresses the skilled labor gap by combining perception, reasoning, and control to deliver risk-aware, reliable, continuously improving field-ready AI. We tackle difficult manufacturing challenges every day, especially in welding, where fast and complex physics demand new approaches. We are hiring two Senior Machine Learning Engineers with complementary focuses: World Models and Reinforcement Learning. We offer daily free lunch, flexible PTO, comprehensive medical, dental, and vision coverage, 6 weeks of fully paid parental leave plus an additional 6-8 weeks for birthing parents, a 401(k) plan through Empower, and generous referral bonuses. We also foster a diverse, inclusive environment where ideas are welcomed and valued.

last updated 33 week of 2026