Research Product Manager, Robotics

WorldEngine; San Francisco (CA); Presencial; USD 150 000 - 200 000; Tempo integral

Resumo da oferta

Frontier AI is seeking a Research Product Manager to lead our research-led go-to-market for robotics data and embodied AI benchmarks. You’ll align researchers with Frontier AI’s data needs, translate conversations into pilot briefs, and hand them to Product to build.

You’ll work closely with our research lead on what we publish and what we keep, shaping the roadmap through hands-on data work in our robotics lab and collaborations with frontier robotics labs.

Qualificações

• Hands-on technical background in engineering or research.
• Proven ability to influence researchers and stakeholders.
• Experience with pre-training or post-training data in labs or vendors.
• Familiar with robot learning models and evaluation methods.
• Track record landing external projects or pilots.
• Comfortable in a startup environment without a playbook.

Responsabilidades

• Build trusted relationships with researchers and spend stakeholders.
• Advise on robotics data for pre-training and post-training.
• Run benchmark and publication roadmap with the research team.
• Co-design pilot prototypes and define requirements with researchers.
• Hands-on data work in robotics lab and QA evaluations.
• Forecast future demand and feed into product roadmap.

Conhecimentos

Engineering background

Strong communicator

Research collaboration

Robot learning models

External projects

Startup mindset

Formação académica

Engineering or research background

Descrição da oferta de emprego

Frontier AI is going physical, and the labs building robot foundation models are bottlenecked on real-world data — but no one agrees yet on what to collect, how to capture it, or how to tell whether it helped. The problems to solve in the industry are wide open.

As Research Product Manager, you’ll run our research-led go-to-market: the benchmarks, publications and post-training results that show labs where their models fall short and what data closes the gap. You’ll work researcher-to-researcher with frontier robotics and embodied AI labs, turn those conversations into pilot briefs for data types that don’t exist yet, and hand them to Product to build. You’ll work directly with our research lead on what we publish and what we keep.

What You'll Do

• Build trusted relationships with researchers and with the stakeholders who control spend at frontier robotics and embodied AI labs, and be their technical point of contact
• Advise customers on robotics data for both pre-training and post-training, from egocentric video, teleoperation, tactile capture to model evaluation.
• Run the benchmark and publication roadmap with our research team
• Co-design pilot prototypes and define requirements with researchers across task and environment mix, capture configuration, annotation and QC criteria, delivery format, and the success metrics
• Get hands-on with the data in our robotics lab from inspecting early samples, playing back data on robotics, and doing quality evaluations yourself
• Capture forward demand, predicting what labs will need in 6-12 months, and feeding it into the company’s roadmap

What We're Looking For

• Hands-on technical background, with a background in engineering or research
• Strong communicator who can influence and build trust amongst researchers
• Prior work with researchers on pre-training or post-training data at a lab, model company, or data vendor
• Fluent in how robot learning models are trained and evaluated today from imitation learning, world models, to VLA-style policies
• Has scoped and landed external projects, pilots, or deals before
• Comfortable operating without a playbook in a new org or startup environment

Nice to Have

• Robotics, autonomous vehicles, or physical AI experience from a lab, graduate program, or industry team
• An existing network among robotics and AI researchers
• 3+ years in product, solutions, consulting, or another customer-facing role at a startup or high-growth company
• Experience designing evaluations or benchmarks for ML systems