Sr. ML Research Scientist
Serve RoboticsFull Time
Senior (5 to 8 years), Expert & Leadership (9+ years)
Candidates must have a PhD or equivalent experience in Machine Learning or Robotics, with a track record of published research in robotics. They should possess practical experience building training VLA models and/or robotics data, along with 3+ years of industry experience in areas such as robotics, computer vision, embodied AI, sim-to-real, imitation learning, reinforcement learning, and vision language actions models. Strong written and verbal communication skills, intellectual curiosity, empathy, and the ability to work autonomously are also essential. Experience with robotics hardware, deploying ML models on robotic systems, and working with teleoperated or human-driven robotics data are considered advantageous.
The Machine Learning Research Engineer will collaborate with robotics customers to advance the use of VLA data and develop ML pipelines for training and fine-tuning models using Scale's data. Responsibilities include conducting research on robotics data collection, cross-embodiment training, and policy fine-tuning, as well as developing novel methods for evaluating VLA models and creating new robotics industry benchmarks. The role also involves partnering with cross-functional stakeholders and customers to improve data collection and collaborating with product teams to integrate ML outcomes into Scale's platform.
AI platform for data and models
Scale AI provides a platform that helps businesses develop AI applications by utilizing their enterprise data to customize generative models. The platform includes tools for collecting, curating, and annotating data, as well as features for evaluating and optimizing models. Scale works with a variety of clients, including major tech companies like Microsoft and Meta, government agencies such as the U.S. Army and Airforce, and startups like Brex and OpenSea. What sets Scale apart from its competitors is its comprehensive suite of tools and services that focus on safely unlocking the value of AI. The company's goal is to enhance the performance of advanced language models and generative models, making AI more accessible and effective for its clients.