Machine Learning Engineer | ML Engineer
Austin, United States
Engineering - Software
Permanent
2026-10-01
We are looking for some of Australia's best machine learning engineers ready to relocate to Austin, Texas, where you will build the neural networks that let self-driving cars see and decide in real traffic. USD 150,000 to 220,000 base (around AUD 220,000 to 320,000) plus bonus, relocation and immigration covered.
We are partnering with an autonomous vehicle company whose driverless cars run in real traffic. They want engineers from Australia, and they are hiring more than one machine learning engineer into a single team covering perception, behaviour prediction and motion planning. Where you land is settled after interview, by what you are strongest at.
To be straight about what this is not: it is not a pipelines, MLOps or LLM integration role. It is neural network work you have trained yourself, deployed on real hardware. If your story is mostly prompts, pipelines and APIs, or simulation only, or research that never shipped, this one is not the fit.
What you will do
Day to day you work in Python with PyTorch, TensorFlow or JAX. Inference runs in C++ on the vehicle, so you will work inside that code too.
- Perception. Train the models that see, and decide which data they need from a fleet that records far more than any person could review.
- Prediction and planning. Model how pedestrians, cyclists and other drivers are about to move, and turn that into how the car should act.
- Evaluation. Build metrics that track real on-road behaviour, because a model that scores better offline can still drive worse.
- Failure analysis. Dig into rare scenes from real driving logs and feed what you find back into the data and the model.
- Fixed compute. Make hard calls between accuracy, latency and memory on embedded hardware, and defend them.
What you will bring
- At least three years training neural networks yourself, from data through training into production or onto real hardware, and keeping them working once there.
- Real depth in one area: computer vision, or behaviour prediction and motion planning, ideally on vehicles, robots, drones or similar physical systems.
- Python and a modern deep learning framework as your everyday tools.
- Working C++, enough to read and work in the code your model runs inside.
- Ownership from prototype to deployment, including working out why a model underperformed and fixing it, and getting your own data at scale with SQL rather than waiting on someone else.
Nice to have
Computer vision work such as detection or segmentation, especially on camera or LiDAR data. Making a model meaningfully faster on target hardware and knowing what you gave up for it. Machine learning for autonomous vehicles or robotics. Published work or open source code they can read. Good judgement on which new architectures, transformers and multimodal models included, actually apply to the problem.
What is on offer
- Salary in US dollars. Mid level USD 150,000 to 180,000 base with a 20% target bonus. Senior up to around USD 220,000 base with about 22% bonus, plus equity. Around AUD 220,000 to 320,000 base across the range.
- Relocation covered. Relocation and US immigration costs, corporate housing on arrival, flights for your family and help with a car.
- Leave and benefits. 21 days leave, 10 sick days and 8 public holidays, health premiums paid, a 401(k) with a 4% match, and 20 weeks primary parental leave after a year.
- Work that is judged on what you built. No advanced degree needed and no prior autonomous vehicle experience needed. Shipped systems count far more than credentials.
- A team that ships together. Onsite in Austin, working directly with the planning, infrastructure and vehicle software engineers who take your model's output.
Who can apply
Open to Australian citizens. Relocation and US immigration costs are covered.
Apply
ScaleUp Recruitment welcomes applicants from every background.
Apply now, or email sam@scaleuprecruitment.com.au for a confidential chat.

