Staff Research Engineer, Hardware Intelligence

About Nominal

Engineers today are building some of the most ambitious machines humanity has ever seen. They should have AI that’s purpose-built for them.

Nominal is building the Hardware Intelligence Platform that gives engineers increased visibility into complex physical systems, helps them make better decisions, and ultimately gets their products into the real world faster. It brings together the systems, data, intelligence, and applications hardware engineering requires to meet the AI moment. Hardware engineering teams use Nominal to simulate, test, manufacture, and operate the physical systems that keep the world moving, from spacecraft and autonomous vehicles to next-generation energy systems and robotics.

Our customers span aerospace, defense, manufacturing, transportation, energy, and robotics. They include Anduril, Shield AI, Wisk Aero, Pratt Miller Motorsports, Antares, and Shinkei Systems, amongst others, alongside programs within the U.S. Air Force and Navy. Our team draws from SpaceX, Palantir, Anduril, Applied Intuition, Tesla, and other leading companies, united by a common mission: giving hardware engineers the tools to build the future with speed, safety, and confidence. We're a fast-moving team that owns problems end-to-end, works across disciplines, and thrives at the intersection of hardware and software.

Some of the world’s leading technology investors, including Sequoia, Lightspeed, Founders Fund, General Catalyst, Lux Capital, and Insight Partners, are backing Nominal’s ambition to change how machines are built and operated. We’re looking for engineers who want to take on hard problems and help make a new generation of world-changing hardware possible.

About Hardware Intelligence

We're going after the biggest opportunities to accelerate how the world engineers: agents that are expert at engineering physical systems, not just competent at calling tools. Nobody has a recipe for that yet. Hardware gives us an edge most domains lack: real test results to check agent work against. You'll set the research agenda, decide where to turn that ground truth into better context, better tools, or models of our own, run the first experiments yourself, and grow the researchers who run the next ones.

What you'll do

• Set the research agenda for Hardware Intelligence: pick the problems, place the bets, and define what winning looks like each quarter.
• Figure out how to store and represent hardware context, so agents reason over telemetry, test campaigns, requirements, designs, and simulations instead of just retrieving them.
• Teach agents to work the tools engineers already use, from CAD to simulation to analysis software, through computer use.
• Connect simulation to the physical world: calibrate simulations against measured results, and use simulations to plan and explain tests.
• Build distributed intelligence: collections of agents that coordinate across programs, fleets, and teams.
• Push intelligence to the edge, with agents and models that run on or near the hardware, on-prem, and air-gapped.
• Build RL environments and verifiers over real engineering work (test campaigns, anomaly investigations, design-versus-measured checks), validate them with hardware experts, and find where they break before we train on them.
• Lead post-training (SFT, RL, distillation) of open-weight models when a model of our own beats a frontier model with better context, and back every call with data.
• Hire and mentor researchers, and publish benchmarks and results worth sharing outside Nominal.

What you'll bring

• 8+ years in ML research or engineering, including a research direction you owned that changed a shipped model or product.
• Deep post-training experience with RL, reward models, or verifiers, and the debugging that comes with them: reward hacking, instability, contamination.
• Statistical rigor: you design experiments that don't fool you, and you know when a metric has stopped meaning anything.
• Research taste: you've picked problems that mattered and killed the ones that didn't.
• The judgment to know when to measure, when to train, and when a better prompt, tool, or context is the answer.
• A track record of setting technical direction across a team and raising the bar for the people around you.
• You build with modern AI coding agents (Claude Code, Cursor, Codex) every day, and stay curious and open to better ways of working. The tools keep changing, and so do we.

Nice to have

• You've led post-training or evals at a frontier lab or an AI-native company.
• You've published benchmarks, environments, or methods that others use.
• You've worked on computer-use agents, multi-agent systems, or knowledge representation for agents.
• You've worked on scientific ML, simulation, digital twins, or time-series foundation models.
• You've trained or served open-weight models on-device, on-prem, or in air-gapped environments.
• You've worked in test, reliability, or verification engineering for physical systems.

Benefits/Perks

• 100% coverage of medical, dental, and vision insurance
• Unlimited PTO and sick leave
• Free lunch, snacks, and coffee
• Professional Development Stipend
• In-office hardware lab with a $250 project stipend
• Annual company retreat
• 12 weeks parental leave
• Fertility and family planning ($15,000 lifetime)
• Wellness stipend ($100/month)
• $500/year hardware sensor stipend
• 401(k)
• Employee-friendly equity terms
• In-office 5 days per week

All qualified applicants will receive consideration for employment without regard to race, color, religion, sex, sexual orientation, gender identity, or national origin.

ITAR Requirements

To conform to U.S. Government export regulations, applicant must be a (i) U.S. citizen or national, (ii) U.S. lawful, permanent resident (aka green card holder), (iii) Refugee under 8 U.S.C. § 1157, or (iv) Asylee under 8 U.S.C. § 1158, or be eligible to obtain the required authorizations from the U.S. Department of State. Learn more about the ITAR here.

Compensation

The base pay range for this role is $268,000 - $325,000 per year.