Software Engineer, Automated Calibration Services

About the role

We are hiring software engineers into Automated Calibration Services, the team that keeps qubits inside spec while programs are running. Physicists design the calibration protocols; you will build the software that monitors, automates and drives them, triggered by a schedule, a compiler request, or a monitoring event.

The device drifts. A calibration valid an hour ago may not hold, and the machine cannot be taken out of service for long. The work is deciding when to run a routine, which one, in what order, and how to establish afterwards whether it worked.

You will own the execution path at room temperature, on the accelerators and at the logical qubit layer: the dependency graph, the realtime service and FPGA engine that run protocols against the device, and the calibration store the rest of the stack schedules against. Control & Error Correction fits calibration into duty cycles, and Compiler Services compiles against the device state you publish.

Based at our Sydney facility, you will work daily with the physicists who design the protocols and the engineers who run the control plane. This is a role for someone who wants their software to drive physical hardware, and who can take a protocol written by a scientist and hand back a routine they trust.

Role responsibilities

• Build the orchestration that runs calibration protocols: the dependency graph, ordering, scheduling, retries and the decision to run at all
• Implement the protocols the physicists hand over as automated, monitored, repeatable software routines
• Build and run the realtime calibration service, and drive the FPGA calibration engine in the mesh
• Own the calibration store: results, provenance, versioning, and the queries the rest of the stack depends on
• Build the drift characterisation and qubit health analytics that decide when a routine needs to run again
• Support every trigger path: scheduled, compiler-requested, and monitoring-driven
• Work with Control & Error Correction so calibration fits inside execution duty cycles instead of interrupting them
• Publish the device state, noise models and duty cycles that Compiler Services compiles against
• Instrument the whole path so a calibration run is reproducible, explainable, and attributable to a device and a protocol version
• Analyse device data to separate drift from measurement noise, and to establish whether a routine improved the qubit
• Turn what physicists know into software they no longer have to run by hand, and keep them in the loop on what it does
• Document the protocols as implemented, the dependency model and the operational procedures

Your experience

Essential

• 4+ years building production software, ideally where the software drives physical hardware
• Strong Python, including the scientific stack (NumPy, SciPy, pandas), plus a systems language such as Rust or C++
• Workflow orchestration and dependency graphs: DAG scheduling, ordering, failure handling and partial re-runs
• Numerical work in practice: curve fitting, parameter estimation, optimisation, and knowing when a fit is not to be believed
• Time series and signal data, including the basics of filtering and characterising noise
• Working with instruments or devices that are stateful, slow, and sometimes unavailable
• Data provenance and reproducibility: knowing which inputs, versions and device state produced a result
• Observability, and measuring a system that runs unattended
• Reading a protocol written by a scientist and turning it into code, then talking it back through with them
• Git workflow, code review, CI, and testing code that depends on hardware
• Clear technical writing

Nice to have

• Laboratory instrument control: QCoDeS, Labber, pyvisa, SCPI, or an in-house equivalent
• Qubit calibration, tune-up or device characterisation, on any qubit modality
• Machine learning applied to device tuning: Bayesian optimisation, Gaussian processes, active learning, or computer vision on device data
• Anomaly and drift detection on production telemetry
• Control theory and closed-loop feedback
• FPGA or HDL exposure, enough to work with the engineers who build the calibration engine
• Realtime or embedded systems, where deadlines are real
• Orchestration frameworks such as Airflow, Dagster, Prefect or Temporal
• Quantum computing exposure of any kind (no physics degree required)
• Publications, open source contributions, or a hardware automation project you are willing to talk through