Onsite_Data Architect- MS Fabric_

Job Title:

Data Architect - MS Fabric

Location:

Columbus, IN

Duration: 6-12+ months

Contract

Seeking a Data Modeling Architect to define and implement the enterprise modeling framework for an advanced data fabric architecture. This role translates business capabilities, domain concepts, and source-system structures into governed conceptual, logical, physical, and semantic models. The architect establishes reusable data contracts and canonical definitions that support interoperability, discoverability, trusted data products, and analytics across Microsoft Fabric and connected enterprise platforms.

The role requires strong data modeling expertise and hands‑on implementation experience with Microsoft Fabric, including One Lake, Lakehouse, Warehouse, Data Pipelines, and Power BI semantic models. The architect works directly with business domain owners and the Data Team to align models with enterprise architecture standards, operational processes, and measurable business outcomes.

• Enterprise Ontology and Domain Modeling
• Define an enterprise ontology and domain taxonomies that establish consistent business terminology, entity definitions, relationships, and business rules across functional domains.
• Develop conceptual and logical models that capture domain boundaries, shared entities, reference data, cardinality, hierarchies, and cross-domain relationships.
• Select appropriate modeling patterns, including normalized entity-relationship models, dimensional models, Data Vault, and canonical data models, based on integration and consumption requirements.
• Define authoritative sources, system-of-record relationships, data ownership, and master and reference data alignment for shared business entities.
• Data Fabric and Physical Implementation
• Design physical models for Microsoft Fabric Lakehouse and Warehouse workloads, including Delta table structures, partitioning strategies, keys, schema evolution, and query-performance considerations.
• Design and govern open table-format implementations using Apache Iceberg, Apache Hudi, and Delta Lake across lakehouse workloads.
• Select the appropriate table format based on interoperability, transaction requirements, ingestion patterns, query-engine compatibility, performance, cost, governance requirements, and operational supportability.
• Define physical table-design standards for partitioning, clustering or data layout, compaction, file sizing, retention, schema evolution, partition evolution, and metadata management.
• Establish patterns for ACID transactions, snapshot isolation, time travel, rollback, auditability, and reproducible data consumption across curated data products.
• Define incremental-processing patterns using change data capture, merge and upsert operations, deletes, deduplication, slowly changing dimensions, and late-arriving data handling.
• Define table-maintenance and optimization processes, including compaction, small-file remediation, orphan-file cleanup, snapshot expiration, retention controls, and performance monitoring.
• Define interoperable data product schemas and data contracts to support governed exchange across ingestion, transformation, serving, and consumption layers.
• Design source-to-target mappings and transformation specifications for ERP, CRM, operational systems, APIs, and other enterprise data sources.
• Apply bronze, silver, and gold layer patterns while defining fit-for-purpose data products and reusable integration assets across One Lake, Lakehouse, and Warehouse.
• Work with engineering teams to implement batch and near-real-time integration patterns, including incremental ingestion, change data capture, and event-driven data flows where appropriate.
• Define standards for schema versioning, backward compatibility, and model lifecycle management as source systems and business requirements evolve.
• Establish cross-platform interoperability standards for data products consumed through Microsoft Fabric, Databricks, Snowflake, Spark, Trino, and compatible query engines.
• Semantic Layer, Metadata, and Governance
• Design governed Power BI and Fabric semantic models with conformed dimensions, consistent measures, calculation logic, and business-friendly terminology.
• Establish metadata requirements for business glossaries, technical catalogs, lineage, classifications, ownership, and policy enforcement using Microsoft Purview or equivalent capabilities.
• Align models with data quality rules, validation controls, observability requirements, and master data management…