bluebill
bluebill
Data and ML
Machine Learning Engineer
Build ML engineering tasks: training and evaluation code with real bugs, and the tests and fixes that prove them.
- Pay
- $110 to $160 AUD an hour
- Hours
- 5 to 20 hours a week while a project runs. Project-based. Work is offered as customer projects open, and you choose when you work.
- Type
- Contract
- Location
- Remote (Australia and international)
- Posted
- 5 Oct 2026
Apply on a laptop or desktop with a camera and microphone.
What you will do
- Create runnable ML projects with realistic bugs: leakage, wrong metrics, shape errors, unstable training.
- Write tests and checks that catch each bug, and a verified fix.
- Explain how the bug would have changed results and how you found it.
- Review other engineers' tasks for realism, reproducibility and test quality.
What you will need
- 3+ years in machine learning engineering or applied ML in production.
- Fluent in Python and PyTorch or a similar framework.
- Clear written English.
Nice to have
- Experience with LLM training or evaluation.
- Research or publication experience.
How you are assessed
ML correctness, test quality, diagnosis, clear explanation.
Applications and work are reviewed against these criteria.
How to apply
A 20 minute video interview: five engineering questions, up to 3 minutes each.