- ML engineering in production: demonstrated experience owning the full Data Science lifecycle — data, training, evaluation, deployment, serving, and ongoing operations — on production systems, not just prototypes. This includes monitoring for data drift, training/serving skew, and model degradation; the production discipline, not just the modelling.
- Data fluency: strong working knowledge of data processing tools — BigQuery (SQL), pandas, Dataflow, or equivalents — and the ability to reason about data quality as a first-class concern.
- ML fundamentals: fluency across supervised and unsupervised learning, ranking models, recommendation systems, evaluation metrics, overfitting, and model selection.
- Cloud platform experience: practical experience with GCP (our primary environment — Vertex AI, BigQuery, Dataflow) and AWS, both of which are in use across our platform.
- Python: strong production-grade Python; comfortable in a codebase where your models and services are consumed by other engineers.
- Technical leadership: experience leading the technical design of ML projects, including articulating trade-offs and driving decisions within a team.
Valued
- Experience with LLM-based systems (RAG, structured generation, evaluation, cost management), particularly on GCP/Vertex AI or similar.
- Experience building or contributing to shared ML infrastructure or internal platforms (feature stores, serving layers, shared pipelines).
- Familiarity with recommendation systems, personalisation, or search ranking at scale.
- Experience working in a cross-craft team alongside software engineers and data scientists.
- Understanding of AB testing methodology and experimentation platforms.
- Familiarity with Java – our model serving layer is built on Java services.
What we are looking for beyond the CV
Smarties is a small team carrying a large portfolio. The ideal candidate is someone who:
- Takes ownership seriously – not just of the model, but of the system, the data, and the outcome.
- Thinks in leverage – prefers a solution that improves ten products at once over ten separate solutions.
- Is honest about uncertainty – raises problems early, shares trade-offs openly, and does not treat "keeping the lights on" as someone else's job.
- Works well across disciplines – can design with a data scientist, review a service with a software engineer, and explain an ML decision to a product manager.
- Cares about data quality – understands that smart features are only as good as the data behind them, and acts accordingly.
This role description summarises the typical functions of the role. It is not an exhaustive list of all possible responsibilities, tasks, or duties.
BENEFITS: WHAT’S IN IT FOR YOU?
Work and life should fit together, so we offer a range of benefits focusing on well-being, development, and team spirit. The final package will depend on the contract type we agree on.
- Standard benefits: life insurance, private health care [Luxmed], Multisport card, lunch vouchers, company share programs, and assistance for everyday worries and serious health problems.
- Learning opportunities: access to the Learnebly platform and LinkedIn Learning, English classes, and a book library,
- Parental support: additional +10 days maternity / +20 days paternity leave, funding for nurseries and kindergartens
- Hybrid model: 2 days/week in the office and flexibility to work from almost any location for up to 30 days a year.
- Office perks: centrally located offices with car and cycling parking, and home office equipment provided.
- Career growth: a clear career path with opportunities to rotate between projects, teams, domains and roles under the guidance of highly skilled senior colleagues
- High engineering culture: unique software engineering culture with a high level of test coverage and agile environment [read about our tech stack and engineering