Posted 4 weeks ago

Staff Software Engineer (System Orchestration)

IonQ

Oxford, England, United Kingdom 🇬🇧

Onsite
Full Time
Airflow
Software Engineering

Apply on source site

We are looking for a highly technical Staff Software Engineer to join our System Orchestration team in System Software at Oxford Ionics
In this role, you will be responsible for building and operating the software that keeps our systems calibrated, decides what they do next and keeps the record of what they did, contributing to the team’s mission of keeping our quantum computers running useful work around the clock
The trapped-ion quantum computers we are building will control 10,000 ions within two years, and all of that has to be kept in tune - settings that drift over minutes, hours and days
Without it a computer does not stop; it quietly starts producing invalid output
Calibration routines, scientists’ experiments and customer workloads all compete for physical hardware, and our team builds the control plane that arbitrates that: it schedules and prioritises programs, hands exclusive access to one at a time and takes it back cooperatively, streams results while a program is still running, and keeps the record of everything that ran
It runs on every quantum computer in our fleet
Its job is to keep that machine busy and productive: running quantum programs rather than waiting in between, and running them in spec - on a machine still hitting the fidelity it is meant to hit
The ambition is for that to run without a person in the loop: machines that assess their own readiness, call for the calibrations they need, and keep a full history of how they were tuned
You’d work closely with the runtime team who own real-time control, the systems teams who build and maintain our quantum computers, and the operators and scientists who run them
This is a hands-on engineering role: you would be writing this software alongside our other staff engineers, not directing others who write it. You’ll play an important part in shaping the services that support our work in building the world’s most advanced quantum systems. Previous exposure to quantum systems is not required
In this position, you will work across the services that make up the control plane, supporting the effective delivery of a system every machine in the fleet depends on
You’ll have the opportunity to remove whole classes of failure and raise the bar for the engineers around you, contributing directly to the success of the System Software function
Take on the problems that span the whole control plane - the durability and recovery contracts the rest of the design rests on
Make a machine’s own assessment of its readiness something the rest of the control plane can rely on.Do the deep diagnosis of failures that only appear on a live system, under real load
Remove whole classes of failure across the fleet, rather than instances of them
Draw the boundaries between our services, settle them in writing, and move the code across them yourself
Make the correctness of a design demonstrable rather than argued
Build the tooling and test infrastructure that let the whole team move faster
Evolve how our services are deployed and operated as the fleet grows
Raise the bar of everyone around you through the code you write and the reviews you give
To be successful, you will need deep systems engineering experience, a track record of operating services where losing data is not an option, and the judgement to tell a problem worth solving structurally from one worth patching. We’re looking for someone who wants to stay hands-on in the code at this level and can contribute effectively within a fast moving, highly technical team
Deep experience building and operating stateful networked services in production, exposed over gRPC or REST, with a real grasp of error handling, observability and failure modes
Expert systems engineering skills. Our System Orchestration layer is written in Rust, so either solid Rust experience or deep proficiency in another systems language and the appetite to work in a Rust codebase, including async programming
A track record of pinning down failures that others could not
Excellent communication skills both within software and across disciplines
A track record with data durability and correctness - persistence design, crash recovery, and migrations against live data you cannot afford to lose. We use PostgreSQL
Depth in the scheduling and workflow landscape - Temporal, Hatchet, Airflow, Slurm, Postgres-backed queues - and the judgement to borrow from it rather than reinvent
A habit of settling design questions in writing, with the options compared as peers
The instinct to reach for determinism when chasing a timing or interleaving bug, rather than for more logging