Technology company building automation and machine learning platforms. The project involves an enterprise data platform built on a data mesh approach, where domain teams own their data as products while a central platform team provides governed, reusable, self-service capabilities. Machine Learning is the next major capability being built into the platform, moving models that currently run outside it into a production-grade, adoptable MLOps environment.
What you will do
- Own the design, implementation and roadmap of the MLOps capability end to end.
- Assess the existing data platform and refine the MLOps architecture to fit its governance and structure.
- Implement the full ML lifecycle on Databricks and MLflow, including experiment tracking, model registry, feature tables, packaging, model serving and monitoring.
- Build controlled promotion of models across development, staging and production using CI/CD practices.
- Build automated retraining, drift and skew detection, along with alerting and rollback mechanisms.
- Productionise both batch and near real-time inference pipelines.
- Provide domain teams with cost visibility and monitoring for ML workloads, treating models as data products with clear ownership and lineage.
Location and work model
