You will take models from data assessment through to production serving. This role covers the full path, including the parts that are not modelling: data quality, baselines, evaluation and monitoring after launch.
We expect you to be willing to tell a client their data will not support what they want.
What you will do
- Run data readiness assessments and report honestly on feasibility
- Build and evaluate models against measured baselines
- Deploy models behind versioned APIs
- Build drift monitoring and retraining pipelines
- Explain model behaviour to non-technical stakeholders
What we are looking for
- Three or more years shipping machine learning to production
- Strong Python, pandas and scikit-learn or PyTorch
- Experience with model evaluation beyond accuracy
- Understanding of data leakage, censoring and selection bias