02 / FINANCIAL SERVICES
Financial Data Engineering & Settlement Reconciliation
Designed and developed scalable Azure data pipelines for financial transactions, settlement records, merchant data, and clearinghouse information. Ingestion, distributed transformation, reconciliation, and validation delivered curated datasets for finance reporting and analytics.
Fiserv · Data Engineer · June 2023 – May 2026THE CONTEXT
A clear starting point.
This work was part of my Data Engineer role at Fiserv, June 2023–May 2026. I translated settlement, reporting, and audit requirements into Azure-based pipelines, source-to-target mappings, and curated financial data models with finance and business teams.
THE CHALLENGE
What needs to work.
Transaction and settlement records needed a consistent basis for comparison. Currency handling, timestamps, merchant classifications, duplicates, and missing values could affect validation. Finance teams needed exceptions they could investigate and reporting datasets they could trust.
THE APPROACH
Inside the decisions.
Prepare comparable financial records
Built batch and near real-time ingestion from relational systems, APIs, and flat files with Azure Data Factory and ADLS Gen2. PySpark and Spark SQL transformations standardized financial records, currencies, timestamps, and merchant categories.
Apply quality and reconciliation controls
Implemented schema, null, duplicate, and referential-integrity checks, then compared internal transaction data with clearinghouse settlement records to identify mismatches.
Connect exceptions to financial reporting
Automated exception reporting and developed SQL logic for settlement amounts, net fees, and transaction-level revenue impact. Curated reporting layers supported settlement summaries, fee analysis, and audit reporting.
Deliver and optimize analytical datasets
Built curated datasets in Azure Synapse Analytics and Azure SQL Database for settlement analysis, finance reporting, and audits. Optimized joins, aggregations, partitioning, and caching in Spark workloads.
Monitor the reconciliation path
Used Azure Data Factory orchestration, retries, operational logs, Azure Monitor, and ETL health checks to surface failures, delayed feeds, and reconciliation breaks.
THE TOOLKIT
OUTCOMES & VALIDATION
Evidence over assumptions.
Automated transaction-to-settlement comparison and exception reporting, replacing manual spreadsheet-based audit workflows. Curated datasets supported finance operations, settlement reporting, and investigation of reconciliation breaks.
A financial discrepancy is useful only when its supporting records and reporting context are clear enough to investigate.