Senior Software Engineer, Embedded Platform

Bonsai Robotics•San Jose, CA

2d•$180,000 - $215,000•Hybrid

About The Position

We are seeking a Senior Software Engineer, Embedded Platform to own the embedded Linux platform that keeps our fleet running in the field. Our machines operate in orchards and row crops far from the office, on intermittent connectivity, and they need to be updated, configured, monitored, and debugged remotely. This role owns that end to end: the Yocto-based operating system image, the over-the-air update path, fleet configuration and provisioning, and the observability infrastructure that tells us what every machine is doing. The ideal candidate is a Linux systems engineer who is comfortable from the bootloader up to a cloud dashboard, and who treats "we had to drive out to the field to fix it" as a bug.

Requirements

Strong computer science fundamentals and systems fluency
Bachelor's, Master's, or PhD in Computer Science, Robotics, Electrical Engineering, or a related field, with 5+ years of industry experience in embedded Linux, infrastructure, or robotics platform engineering
Strong proficiency in Rust, C/C++, and Python
Linux systems programming — networking, IPC, process management, systemd, device drivers
Experience building or operating embedded Linux distributions with Yocto/OpenEmbedded, Buildroot, or a custom BSP, including kernel and device tree work
Hands-on experience with OTA update systems (Mender, RAUC, SWUpdate, OSTree, balena, or equivalent) including A/B partitioning, atomic updates, and rollback
Fleet device management at scale — provisioning, remote configuration, secrets distribution, and remote access to devices behind NAT or on cellular links
Observability and monitoring infrastructure (Prometheus, Grafana, OpenTelemetry, or similar) applied to intermittently connected edge devices
CI/CD, containerization, and cross-compilation for ARM64 embedded targets
Cloud infrastructure experience (AWS or equivalent) and infrastructure-as-code
Strong debugging instincts across the whole stack, from a kernel log on a machine in a field to a cloud metrics pipeline, with the ability to work independently and collaboratively

Nice To Haves

Embedded firmware programming — bare metal or RTOS development, and hardware bring-up on custom boards
CAN bus systems for off-road and agricultural equipment — J1939, UDS/ISO-TP diagnostics, and firmware flashing to CAN-connected ECUs
Experience shipping software to fleets operating on unreliable or low-bandwidth connectivity
NVIDIA Jetson/Orin platforms and L4T (Linux for Tegra, NVIDIA's Jetson board support package)
Secure boot, image signing, and key management (SOPS, AWS KMS, or similar)
Deep understanding of ROS/ROS2 or similar robotics middleware, and familiarity with Zenoh
Exposure to agricultural, construction, or off-road robotics
Experience building developer tooling and reproducible development environments for embedded teams
Data infrastructure experience — high-volume ingest, time-series storage, log pipelines
Familiarity with functional safety processes for off-road machinery (ISO 25119, ISO 13849)

Responsibilities

Own over-the-air (OTA) updates end to end across the fleet — image build and signing, staged rollout, rollback and failure recovery, and update delivery to both Linux compute modules and CAN-connected embedded controllers.
Own fleet management infrastructure — device identity and provisioning, hierarchical configuration and secrets management, remote access, and an accurate live picture of what software every machine is running.
Build the observability stack for a distributed fleet — on-robot telemetry, metrics, logs, and health signals flowing to the cloud, with dashboards and alerting that surface failures before a customer calls.
Own the OS layer — Yocto-based BSP maintenance, kernel and device tree configuration, system services, boot and recovery paths, and reproducible image builds across multiple compute platforms.
Design and maintain build and release infrastructure for embedded targets — cross-compilation, CI/CD, artifact management, and release pipelines that make deploying to a live machine routine rather than an event.
Partner with the firmware, autonomy, and field operations teams to define integration and release requirements across multiple robot form factors.
As an early platform engineer at a growing team, expect to work outside this scope regularly — bringing up new hardware, unblocking a field deployment, or building whatever tooling the team needs next.