Senior Machine Learning Engineer

Machine Learning Engineer – RL & Multimodal Post-Training (Physical AI) OH (onsite) or Remote, US | NY/SF-level compensation | Visa support available

A venture-backed physical AI company is hiring ML Engineers to its Robot Learning team. Unlike most robotics companies, this one is already past the prototype stage: 50+ autonomous systems are live in production with paying industrial customers, the company is approaching 250 people, and it made 80 hires this year.

The team builds foundation models that let industrial robots see, reason and adapt in real time. You’ll work where generative modelling, reinforcement learning and real-world deployment meet, and your models will run on a live fleet rather than only in a paper.

* Use RL-based post-training (RLHF, GRPO, DPO and diffusion-based methods) to train and refine vision-language-action (VLA) and vision-language (VLM) models

* Train on multi-modal data (image, video, force/torque) to improve perception and decision-making

* Design the whole model lifecycle, including data collection, fleet learning and continuous learning from production data

* Build, optimise and ship models that run across a fleet of deployed systems

* Work closely with perception, controls, hardware and software teams to take research into production

* Industry experience training and deploying RL or vision/multimodal models in production. This matters more than academic or postdoc experience.

* Hands-on work on RLHF, GRPO, DPO or other post-training pipelines for vision or multimodal models

* Experience training or fine-tuning VLMs or VLAs at scale

* Strong Python and PyTorch skills; C++ is a plus

* MS or PhD in CS, Robotics, EE, ME or a related field

* Publications at top venues (NeurIPS, ICLR, ICML, CoRL, RLC) are a plus

You don’t need a robotics background. Strong candidates often come from big-tech research labs, generative AI companies and autonomous vehicles. Robotics or simulation experience (Isaac Sim, MuJoCo) is a bonus.