London or Bristol, 3 days in the office, 2 days WFH
At Fractile, we’re building what we believe will be the world’s fastest AI inference chip from the ground up. We’re balanced across hardware and software engineering, and HW/SW co-design is how we work. We move fast, and we help each other move fast. We care about each other, the software we ship, and the people who rely on it.
We build the control chain that powers on, monitors, updates, and protects devices and racks across bare metal, RTOS, and embedded Linux. It’s production-critical software that keeps racks stable, updates safe, and hardware secure. It’s a critical layer in turning a tokens per second benchmark into a tokens per month system, turning great silicon into reliable output at rack scale.
You’ll be there for the first racks coming to life and rollout days where update safety matters. Your work makes the difference between a bad failure and a clean recovery path. This is the work that makes the system something operators can trust.
What you’ll do- Develop firmware and services for management controllers at every level: rack management, BMC, and board/device controllers (RTOS and bare metal)
- Build systems software for board and rack management, from early bring-up through to production deployment
- Implement the critical flows that keep fleets healthy: power-on and provisioning, monitoring/telemetry, updates, recovery, and security
- Integrate with the wider host/device stack (drivers and runtime libraries) so the system works end-to-end
- Work closely with hardware and device software teams to define reliable interfaces, validate behaviour, and debug boundary issues
- Proven experience shipping production-critical, reliable embedded systems across two or more of: bare metal, RTOS, embedded Linux (or deep expertise in one with clear ability to ramp the others)
- Strong C and/or Rust, and excellent engineering judgement around safety, testing, debugging, and failure modes
- A reliability-first mindset: you design for observability, diagnosability, and recovery paths that are predictable and testable
- Comfortable working close to the hardware/software boundary (bootloaders/BSP/kernel-adjacent work) and collaborating tightly with HW, driver, and runtime teams
- Collaborative and high-ownership. You communicate clearly, move fast, and enjoy working through hard problems with others
- Computer Science, Electronic Engineering, Maths, Physics, or related degree and 3+ years of industry experience
- Embedded platform / systems firmware experience (board management, BMC-style stacks, rack controllers)
- OpenBMC experience and/or embedded Linux image build/customisation experience
- Zephyr (or similar RTOS) experience
- Security-minded, with practical instincts around secure boot, signed updates, device identity, and designing systems that fail safe and recover cleanly (direct experience in some areas is a plus)
- Kernel device work (drivers, device tree, bring-up, perf/trace)
- Hardware bring-up experience, including lab tools (logic analysers, oscilloscopes, waveform viewers)
- RISC-V familiarity
- Experience with GPUs or ML accelerators
If you want to build the control software that keeps rack-scale systems stable, safe, and secure for next-gen AI, come build it together.
