Senior Machine Learning Engineer (Reinforcement Learning/World Model)

The Robot Age; Northern (KY); Hybrid; USD 120,000 - 190,000; Full time

Job summary

Path Robotics is seeking a senior machine learning engineer focused on reinforcement learning and world-models for robotic welding. This role centers on embodied AI approaches that operate reliably in variable production environments and transfer learning from simulation to real manufacturing settings.

You will collaborate with perception, reasoning, and control teams to push robust, real-world robot behavior while handling the challenges of physical process variability.

Qualifications

• Strong ML background with reinforcement learning, world models, robotics, or embodied AI experience.
• Ability to build systems that hold up outside controlled lab conditions.
• Comfort with hard physical processes where simulation and reality do not perfectly match.

Responsibilities

• Develop reinforcement-learning and world-model approaches for robotic welding.
• Improve how systems reason about fast, difficult-to-model manufacturing processes.
• Turn real-world robot experience into better behavior and reliability.
• Work with perception, reasoning, and control constraints in production settings.

Skills

Reinforcement learning

World models

Robotics

Embodied AI

Robustness in real world

Job description

What the role actually is

Path Robotics is hiring a senior machine learning engineer for reinforcement learning and world-model work. The role is tied to robotic welding, where the system has to handle complex physical processes and variability on production floors.

The posting frames the work as embodied intelligence for manufacturing rather than generic ML infrastructure.

What you would work on

• Develop reinforcement-learning and world-model approaches for robotic welding
• Improve how Path's systems reason about fast, difficult-to-model manufacturing processes
• Turn real-world robot experience into better behavior and reliability
• Work with perception, reasoning, and control constraints in production settings

What they are asking for

• Strong ML background with reinforcement learning, world models, robotics, or embodied AI experience
• Ability to build systems that hold up outside controlled lab conditions
• Comfort with hard physical processes where simulation and reality do not perfectly match
• Ability to work remotely in the U.S. or from Columbus

Why this one is worth a look

Robotic welding is a demanding proving ground for embodied AI because the process is high-stakes, variable, and physically messy. Path's role is interesting because the learning work is pointed at a real manufacturing bottleneck.