Position: Sr. Enterprise Data Architect JR1044491
Location: New York City, NY (3 Days a week onsite from office)
Client: Mizuho
Domain: Banking / Capital Markets
Job Description:
About the Role
Synechron is seeking an experienced Enterprise Data Architect to define target-state architecture, reusable architecture patterns and migration roadmaps for a complex financial-services data environment.
The role will focus on interpreting stored-procedure landscapes, establishing clear ownership boundaries across domains and medallion layers, and designing a scalable data-product architecture.
You will define reusable patterns for operational applications, ingestion, market and reference data, extracts, enterprise models, master data management, sub-ledger and risk assets.
The successful candidate will combine deep expertise in data mesh, data-product architecture, lakehouse platforms and financial-services data with the ability to produce practical architecture artefacts that can guide implementation.
You will work across business, data, platform and engineering teams to ensure that the platform enables domain ownership without owning the underlying domain data.
Key Responsibilities:
Interpret the stored-procedure landscape analysis architecturally: where a single procedure conflates responsibilities belonging to different domains or different medallion layers, define how it is split.
Define one reusable target pattern per MCDB feature class: operational UI / applications (app), raw ingestion from systems of record (stage), third-party market and reference data (refdata), consumer extracts (extract), enterprise model, and the cross-cutting MDM, sub-ledger and risk assets.
Map every inventory item to a pattern and produce the pattern-assignment matrix.
Compose the target-state conceptual architecture on DEAL: source-aligned data products, aggregate and consumer-aligned data products, the federated governance plane, the self-serve platform and the catalogue of catalogues — with the platform provisioning capability but never owning domain data.
Define the medallion mapping for each pattern
Specify output-port design so that contract-governed subscriptions replace point-to-point extracts, and define the change-feedback loop from consumer to producer.
Produce the complexity, risk and dependency assessment and sequence the roadmap against it.
Deliverables owned
Architecture pattern catalogue, one per feature class (Word; TOGAF pattern cards)
Pattern-assignment matrix covering every inventory item (Excel)
Target-state conceptual architecture (C4 and/or ArchiMate diagram + narrative)
Complexity, risk and dependency assessment with sequenced roadmap (Word + diagram)
Required Skills and Experience:
10+ years in data architecture, 5+ at enterprise or principal level in financial services.
Genuine data mesh / data-product architecture delivery, not familiarity with the literature. Candidates must be able to describe a specific domain decomposition they authored, the contested boundaries within it, and how the platform-versus-domain responsibility split was enforced.
Deep Lakehouse and medallion architecture: Databricks with Unity Catalog, Delta, Snowflake, Kafka, Airflow, Azure Data Lake Storage Gen2. Must be able to reason about equivalents on AWS (Lake Formation, Glue Data Catalog, Amazon DataZone) when comparing options for Mizuho.
Reference and master data architecture: vendor onboarding, identifier crosswalks across FIGI, LEI/GLEIF, CUSIP/ISIN, SEDOL and MIC, mastering and distribution to consuming domains.
SQL Server and T-SQL depth sufficient to reason about stored-procedure decomposition — the ownership boundaries in MCDB are inside the procedures, not the schema names.
TOGAF 9 or 10 certified, with demonstrable use of TOGAF to define and govern patterns; ArchiMate or C4 modelling in production use.
Preferred qualification
FIBO applied to a taxonomy or ontology exercise; ODCS or comparable data-contract specification.
Lineage and catalogue tooling in anger: Solidatus, Unity Catalog, Microsoft Purview, Collibra, Alation, Securiti.ai, Anomalo.
Experience with automated code-intelligence accelerators for SQL/SAS estate analysis.
