Machine learning engineer
Develop machine learning systems that connect modelling choices with repeatable software and data processes.
Remote
A guide to the work this role can cover. Current candidates and availability are confirmed against your brief.
The work to cover
- Build training pipelines, feature processing and inference components.
- Evaluate model behaviour and maintain reproducible experiments.
What to assess
Use the brief to decide which technical evidence matters. An assessment could explore the following:
Review an experiment and explain how data leakage, serving conditions or changing inputs could affect its results.
Consultant profile
Sarah L.Machine learning engineerJohannesburg, South Africa
- How they can help
- Builds training and inference pipelines with monitoring and rollback plans for changing models and operational data.
- Experience
- 9 years
- Qualification
- MSc, Applied Mathematics
- Key systems
- Python / PyTorch / MLflow
Moving a validated model into a working service
The example brief
A data science team has a promising model that must run reliably on new operational data.
What the contribution could look like
- Build repeatable training and inference pipelines with versioned data and model artefacts.
- Define monitoring and rollback behaviour for changing inputs, model performance and service failures.
Working alongside
Data scientists, data engineers, platform teams and the model owner.
Brief us on this role
Start with this role, then tell us what the work needs. You can clear the role and describe the work in your own words.