Senior ML Platform Engineer I - Financial Crime at Wise
United Kingdom
Posted
on-site · full-time · Visa sponsorship
Responsibilities:
- Design and build a declarative, config-driven training pipeline for reproducible model training and deployment.
- Implement model packaging and serving abstractions supporting multiple model types (classical ML, deep learning, etc.).
- Create a model evaluation framework with standard metrics, reproducible comparisons and validation gates.
- Build model monitoring (drift detection, performance alerts, retraining triggers) and maintain audit trails for regulatory compliance.
- Integrate FinCrime lifecycle tooling with Wise’s central ML platform and define clean infrastructure boundaries.
- Enable data scientists by removing operational overhead and increasing productivity across the ML lifecycle.
Requirements:
- Proven production experience building ML platform infrastructure (training, serving, monitoring).
- Strong software engineering skills in Python and experience with Kotlin or Java; solid SQL knowledge.
- Familiarity with orchestration tools (Airflow, Kubeflow), model registries (MLflow) and container-based deployment.
- Experience with model serving, evaluation pipelines, monitoring and drift detection strategies.
- Ability to design internal platforms for adoption and collaborate with data scientists and central platform teams.
- Experience in regulated environments or financial crime detection is highly valuable.