Gradient Robotics is building intelligent robots for data centers and factories. We believe the fastest practical path to capable general-purpose robotics is through industry deployment. We're here to change the world with robotics.
We own and build our full robotics stack: designing the hardware, building the software systems, training ML models, and manufacturing through a global supply chain, all for world-class iteration speed, reliability, and scale.
We've built 3 generations of robots in 4 months with a team of 7, backed by top-tier investors and working toward pilots with the world's largest data center builders. Our team created the best-selling open-source humanoid robots in the US at K-Scale Labs and shipped foundation software at Tesla Optimus and Google.
The RoleThis is a dedicated track for engineers and researchers with extraordinary records who want to stay in the US long-term. Gradient Robotics will file an O-1A petition with top immigration firms on your behalf. Our founders have been through the process ourselves.
These positions require extraordinary technical capability: novel mechanism design, frontier robot learning, or systems performance at the edge of what's currently possible. Founding engineers here make decisions the company will live with for years.
You'll join as a founding engineer in one of three focus areas:
Mechanical. Own the robot end-to-end: analyze payload capacity and kinematics, design mechanisms that are nimble and safe, and take ideas from multi-physics simulation to the last deburred sheet-metal edge.
ML. Train and deploy the models that make our robots useful in the real world. Work across the full pipeline: data collection, VLA pre-training, fine-tuning, and RL-based self-improvement on real hardware.
Software. Work close to the hardware, from kernel and firmware up to the autonomy stack. Own data pipelines that move hundreds of megabytes at single-digit millisecond latency.
Across every track, your work ships to real robots daily.
What You'll DoMake hard design decisions the company will live with for weeks, sometimes months
Ship code, mechanisms, or models to real robots running in real environments
Move fast, iterate hard, and watch your work compound across robot generations
Strength in one of these focus areas:
Mechanical
Mechanical Design: Precision mechanisms, structures, actuators, DFMA
Analysis: Multi-physics simulation, FEA, tolerance stacks, vibration/thermal reasoning
Build + Iteration: Prototyping, test rigs, failure analysis, rapid CAD-to-hardware cycles
ML
Deep Learning: PyTorch or JAX, distributed training, mixed precision
Robot Learning: Imitation learning, diffusion policies, flow matching, or VLA architectures (π₀, OpenVLA, ACT, RT-X, Octo, etc.)
Reinforcement Learning: Sim-to-real transfer, distillation
Simulation: MuJoCo, Isaac Sim, or Genesis
Python: Research-grade code that ships to production
Software
Software: Rust, Python, C++, operating systems, multithreading
Infrastructure: Bazel, Nix, HIL testing, CI/CD
Firmware / Platform: Linux kernel hacking, embedded systems
USCIS requires evidence that you're among the small percentage at the top of your field. Strong candidates typically have several of:
Top-tier publications. CoRL, RSS, ICRA, NeurIPS, ICML, ICLR for ML and learning. Equivalent venues for systems, controls, or hardware.
Patents. Granted or pending, with you as an inventor.
Awards and selections. Best paper, fellowships, competitive prizes, selective programs.
Original technical contributions with measurable impact. Open-source projects with significant adoption, production systems shipped at scale, products with real users.
Press or media coverage. Coverage of your work in mainstream or trade publications.
Peer review or judging. Conference PCs, journal reviewing, hackathon or competition judges.
External recognition. High citation counts, invited talks, keynote speaking.
You do not need every category. You do need a credible case. We work with top immigration counsel to assemble the strongest petition possible.
LogisticsIn-office in San Francisco, CA
This role is structured around O-1A classification. Candidates should hold US work authorization with at least 1 year remaining (F-1 OPT, STEM OPT, H-1B, TN, E-3, L-1, or O-1).
Available to start imeediately.
If that excites you, apply below.
Skills Required
- Extraordinary technical capability in mechanism design, frontier robot learning, or systems performance
- Mechanical design: precision mechanisms, actuators, DFMA
- Multi-physics simulation, FEA, tolerance stacks, vibration/thermal analysis
- Prototyping, test rigs, failure analysis, rapid CAD-to-hardware iteration
- Deep learning experience with PyTorch or JAX
- Distributed training and mixed-precision training experience
- Robot learning experience (imitation learning, diffusion policies, VLA architectures, flow matching)
- Reinforcement learning and sim-to-real transfer experience
- Simulation experience (MuJoCo, Isaac Sim, or Genesis)
- Production-grade Python (research-to-production) skills
- Systems/software skills: Rust, C++, operating systems, multithreading
- Infrastructure experience: Bazel, Nix, HIL testing, CI/CD
- Firmware/platform skills: Linux kernel hacking, embedded systems
- Evidence of extraordinary achievements (top-tier publications, patents, awards, notable OSS or production impact)
- In-office work in San Francisco, CA
- Must hold US work authorization with at least 1 year remaining (F-1 OPT, STEM OPT, H-1B, TN, E-3, L-1, or O-1)
- Available to start immediately
What We Do
The datacenter buildout is the largest industrial project in human history. Gradient builds the autonomous robots that make it possible. Partnered with the world's largest AI infrastructure companies and backed by the industry's best investors, we move fast and build full-stack systems that matter.
