Research Engineer

Research Engineer

About Croutch & Associates

Croutch & Associates partners with high-growth startups, venture-backed technology companies, and innovative organizations to help them build world-class teams. We combine recruiting expertise, operational excellence, and AI-powered workflows to deliver exceptional hiring outcomes for our clients.

Our client is an early-stage, well-funded robotics company building general-purpose robots to automate physical labor, beginning with bi-manual manipulators for industrial applications. Industrial labor represents nearly $1 trillion in annual U.S. spending, with hundreds of thousands of unfilled positions today. The company approaches robotics as a full-stack challenge, combining cutting-edge hardware, machine learning, and deployment infrastructure to achieve 99.9%+ real-world reliability. Their small, highly technical team is advancing large-scale pretraining, reinforcement learning (RL) post-training, and the infrastructure required to scale both.

About the Role

Research velocity, measured by experiments shipped each week, is the primary driver of progress toward building general-purpose robots. As a Research Engineer, you'll work directly with world-class researchers on large-scale pretraining and reinforcement learning (RL) post-training while owning the critical path from GPU training clusters to robots operating successfully in production environments.

What You'll Do

• Implement model architectures and training strategies alongside researchers while optimizing performance at scale.
• Build and maintain robust data pipelines for collection, curation, filtering, and augmentation across vision, proprioception, robotic actions, and language.
• Own training infrastructure, including distributed training, checkpointing, profiling, and debugging across GPU clusters.
• Develop evaluation systems that detect regressions and generate meaningful offline and real-world robotic performance insights.
• Manage the end-to-end deployment pipeline, taking models from training clusters to robots in the field while incorporating production data to continuously improve performance.

What We're Looking For

• Strong software engineering fundamentals with deep machine learning expertise.
• Experience implementing machine learning models end-to-end, not simply using existing frameworks.

Explore our comprehensive directory of visa sponsorship jobs from employers worldwide who are ready to sponsor talented international professionals.

• Hands-on experience with PyTorch or JAX.
• Experience with distributed machine learning training.
• Ability to debug issues across the full technology stack.
• Comfortable working in fast-paced, highly ambiguous startup environments.

Preferred Qualifications

• Experience with large-scale distributed training infrastructure across multi-node, multi-GPU environments.
• Experience working with large multimodal models.
• Publications at leading ML or robotics conferences such as NeurIPS, ICML, ICLR, CoRL, RSS, ICRA, or similar.

Interested in opportunities specifically in United State? Discover our dedicated Visa Sponsorship Jobs in United State page featuring roles from top employers in this location.

• Experience deploying machine learning models to physical robotic hardware with a focus on latency optimization and edge computing.

Compensation & Benefits

• Base salary: $250,000-$400,000, depending on experience.
• Competitive equity package.
• Full-time position.
• Five days per week on-site in San Francisco.
• Visa sponsorship available for qualified candidates who can start quickly.
• Opportunity to join a rapidly growing engineering team building next-generation robotics technology.