The Data Engineer / ETL Specialist builds and runs the pipelines that move data into training, features and continuous validation. Role 2.5 owns catalogue policy and ownership. Role 2.1 consumes contracts and features. Role 2.2 consumes validation feeds. This role makes lineage, quality gates, sensitive-data protection and configuration of datasets/pipelines actually execute.

Standing duties

• Build pipelines only for sources and purposes Role 2.5 has classified and that match the last signed intended purpose. A new source population is ACT02, not a connector.
• Capture lineage and quality at run time so Article 10 evidence is produced by the pipeline, not reconstructed after an incident.
• Feed Role 2.2 with the data Role 2.1 specified for drift and performance monitoring. A pipeline that cannot support validation is an IMP defect.
• Treat datasets, transformation code and feature stores as configuration items with Role 11.1 — no silent schema or source swap.
• Do not treat a quality dashboard or a lineage screenshot as residual-risk acceptance.

Accountable to: CDO / Head of AI Development (or GM: Data Engineering)

Primary domain: Implement (pipelines). Operates pipeline runs that feed validation. Does not Direct or accept residual risk.

Linked EN 18286 clauses: 8.5, 10.1.1, 10.1.1.2

Format: 3 hours of training, EUR 200.