Robotics Software Lead

Robotics Software Lead
GRAM, Los Angeles moving to the San Francisco Bay Area
About the company
GRAM is building a standalone, legged robot for critical infrastructure, with utilities as the initial focus. The locomotion policy is trained in simulation, and the engineering team today is hardware heavy and software light: mechanical and electromagnetics, electrical and power distribution, prototyping and chassis, and engineering leadership.
The role
You will be the architect of the full onboard software stack and the central node of the system. You own the message passing architecture, the middleware and the real time performance, so that every module (perception, controls, planning and drivers) communicates at the right speed with no latency surprises. You are not expected to own locomotion, perception or controls directly.
This is a hands on role, you report to the Head of Engineering, start as an individual contributor, and have a path to build a team as the company scales.
What you will do
Own the onboard software architecture: how perception, estimation, planning and control fit together and communicate.
Own real time performance across the stack, with latency and jitter budgets for every loop that touches hardware, and find where cycle time is being lost.
Define and own the middleware layer and time synchronization across sensors and compute. The team is leaning towards Zenoh over DDS and is migrating its middleware.
Partition workloads across CPU, GPU and microcontrollers so safety critical loops stay isolated from autonomy and AI.
Lead the upgrade of the current ROS stack to a supported release and define a modular architecture across all software layers.
Build the logging and replay tooling used to debug field runs.
What you bring
Real production experience with pub/sub middleware, Zenoh or DDS.
8+ years experience
Real time OS experience, including profiling, tracing and latency analysis.
A solid understanding of concurrency and inter process communication, including memory management.
Strong C++ and working knowledge of Python. Rust is a bonus.
You have designed the software architecture of a robot or a similar real time system and carried it through to production or field deployment.
Nice to have
Experience with legged, humanoid or other mobile robots, ideally standalone and battery powered rather than fixed arm.
EtherCAT or other fieldbus implementations.
GPU platforms such as NVIDIA Jetson and onboard AI inference runtimes.
A mechatronics to software background with real time systems deployed in production.
Experience from autonomous vehicles, aerospace or drones where real time architecture is a core discipline.
Why join
A clean slate architecture, with the freedom to build things properly from scratch.
Early equity, included in the package.
The chance to own a domain and build a team around it.
A small, hands on team where everyone does real work
Interview process
Recruiter screen
Hiring manager screen with the Head of Engineering.
One to two technical rounds, including a live coding exercise and an architecture discussion.
Onsite Interview