Cell Platform Lead

The Role

We are hiring the lead for our Cell Platform team. The cell platform is everything between firmware and application software on the robot cell, plus the electronic hardware it runs on: sensors and cameras, computes and carrier boards, networking, the OS and BSP layer, middleware and data transport, provisioning, and fleet infrastructure.

We treat the platform as an internal product. Its customers are our robotics, software, and manufacturing teams; its contract is a consistent, reproducible, performant base cell that they build on without worrying about the layers below. You will own that product: roadmap, releases, and the relationships with the teams that depend on it.

This is a player-coach role. You will be hands-on-keyboard from day one while defining the roadmap and hiring the engineers who will form the team around you.

What We're Looking For

You have built and shipped platforms on fleets of robots deployed into production for real customers. You know what lab prototypes and pilots don't teach: surviving OTA updates on remote hardware, debugging failures without physical access, and maintaining reliability against customer expectations rather than internal deadlines. Several years of this production-fleet ownership is our hard requirement; typically this looks like 8+ years of industry experience at AV, humanoid, AMR, or drone companies, owning the layer between BSP and autonomy. Skills

Embedded Linux/BSP depth: kernel, drivers, device trees, boot, and BSP management; platform bring-up on NVIDIA Jetson and x86 targets.

Sensor integration and evaluation: has integrated a multitude of sensor types at the driver, SDK, and middleware-wrapper levels, and knows how to design and run sensor evaluations (latency, sync, driver stability, thermal). Emphasis on GMSL2 and PoE cameras.

Real-time and determinism: configures and validates real-time Linux kernels, CPU isolation, IRQ affinity, and RT scheduling policies to guarantee loop jitter budgets for control at 500 Hz–1 kHz; understands the interaction between RT configuration, DDS middleware, and zero-copy transport paths.

Middleware fluency: ROS 2 and DDS/RMW internals, QoS, node composition, zero-copy transport.

Data transport architecture: reasons about functional and optimal data transport on single-compute and distributed-compute architectures from first principles (shared memory, zero-copy, DMA/NVMM paths, network protocols), independent of any one middleware, and evaluates alternatives against measured budgets.

Hardware evaluation and qualification: has sourced, brought up, and qualified sensors/computes/networking on real robots; comfortable owning a BOM and working the seam with mechanical/electrical engineering.