Situation

The client ran fixed income credit analysis and portfolio management across fragmented systems, where security-issuers were identified only by vendor-specific tickers and analyst coverage, security mappings, and index data lived in disconnected, manually maintained tables. This gap held back the company's strategy of becoming a data-driven, AI-ready organization: reconciling security-issuers across source systems (SimCorp Dimension, Bloomberg, index providers), historizing coverage, and linking securities to their security-issuers were all unreliable.

Task

Build a central, integrated data foundation that fulfills the company's data strategy and provides the trusted, AI-ready basis for future ML use cases. It had to deliver a company-owned inventory of security-issuers, resolve security-issuer and security identifiers across heterogeneous sources, and model analysts, coverage, securities, and index data with full historization — migrating existing draft Snowflake SQL into declarative, testable dbt models while routing ambiguous matches to manual review instead of failing pipelines.

Action

• Built a state-of-the-art dbt project on Snowflake, migrating imperative draft SQL into a declarative, CORE-first design.

• Delivered a normalized master-data model of unique business entities plus a galaxy (fact-and-dimension) marts layer for fast analytics.

• Created a company-owned inventory of security-issuers, decoupling issuer identity from vendor tickers via dedicated identifier and reference models.

• Implemented cross-system identifier resolution (vendor tickers, SimCorp Dimension partners, ISIN/CUSIP/FIGI), routing ambiguous matches to an exception model.

• Modeled analysts, coverage, and a canonical security master (iBoxx, ICE ML, SimCorp Dimension) with full historization.

• Applied deterministic keys, dbt tests, and reconciliation analyses, reusable across DEV → UAT → PROD with CI/CD and Dagster.

Result

Delivered a repeatable, test-covered data foundation consolidating security-issuers, identifiers, references, analysts, coverage, securities, and mappings into governed Snowflake tables — pairing a normalized master-data model with a galaxy marts layer for fast analytics. It gives the client an unambiguous, fully historized single source of truth for fixed income credit analysis, automating security-issuer and security matching, isolating data-quality exceptions for controlled correction, and laying the trusted foundation for downstream applications and future ML-driven use cases across DEV, UAT, and PROD.