Data Modeller

Role & responsibilities

- Design, develop, and maintain Conceptual, Logical, and Physical data models for healthcare data warehouses, data marts, and lakehouses.

• Produce each modelling layer as a reviewable, standalone artifact - a conceptual model with clean definitions, a platform-agnostic logical model with explicit grain, and a physical model committed to concrete types, clustering, and materialization.
• Build and maintain complex data diagrams and enterprise entity-relationship maps that stakeholders can actually read and trust.
• Spec and design end-to-end data products from the dashboard / self-serve semantic layer at the top, down through business-object mapping, into data models and the physical store.
• Design semantic / metrics layers that plug cleanly into BI tools (Power BI, Tableau) so analysts and business users can self-serve without re-deriving logic.
• Translate reporting and KPI requirements into durable model structures rather than one-off pulls.
• Forward / Reverse engineering: generate physical SQL- DDL from logical models, and reverse-engineer existing (often legacy) healthcare and commercial databases into clean visual schemas.
• Query optimization: Advanced ANSI SQL and analytical queries to parse hierarchical and longitudinal healthcare data (patient journeys, claims histories, HCP/HCO hierarchies).
• Performance tuning: Indexing strategies, partitioning, clustering keys, distribution styles, and execution-plan analysis to keep queries fast on very large patient / claims / sales datasets.
• Lakehouse paradigm: Model for decoupled storage and compute with practical Delta Lake / lakehouse experience (medallion architecture, incremental models).
• Author and maintain Source-to-Target Mapping (STTM) documents, data definitions, and metadata dictionaries.
• Define and enforce data lineage, data-quality KPIs, and reconciliation logic so the model is auditable end-to-end.
• Create data structures for US pharma domains — Electronic Health Records (EHR/EMR), clinical performance metrics, and/or pharma commercial data (sales, patient, market access, claims, patients).
• Model the messy realities of the domain: HCP/HCO master data, longitudinal patient/claims data, specialty pharmacy feeds, and affiliation hierarchies.
• Bake privacy-by-design into the models to support HIPAA compliance and general data-security guidelines (PHI/PII handling, masking, role-based access, de-identification).
• Work shoulder-to-shoulder with data engineers, clinical/commercial stakeholders, and BI analysts to optimize pipelines and make sure reporting requirements are genuinely met — not just technically satisfied.