Data Platform Modernisation for a FinTech Lender
Three legacy silos consolidated into one real-time intelligence platform
Key outcomes
- Reporting latency reduction
- 50%
- Data silos consolidated
- 3→1
- Executive dashboard delivery
- Real-time
- Regulatory reports on time
- 100%
A fintech lender had data distributed across three disconnected legacy systems — a core banking platform, a CRM and a loan origination system. Reporting required manual exports, took two days and was consistently unreliable. We rebuilt the entire data stack on BigQuery.
Stack
Leadership could not trust their data. Finance, risk and operations teams each had different numbers for the same metric. Month-end reporting was a two-day manual exercise and regulatory reports were frequently late.
How we tackled it, step by step.
Mapped all data sources, schemas and business logic across the three legacy systems
Built event-driven ELT pipelines using Fivetran and custom Python connectors
Designed a layered BigQuery warehouse — raw, staging and mart layers using dbt
Implemented row-level security in BigQuery aligned to organisational roles
Built live executive dashboards in Looker with drill-down to transaction level
Delivered a data quality framework with automated alerting on anomalies and missing data
Outcomes that speak for themselves.
Reporting latency reduction
Data silos consolidated
Executive dashboard delivery
Regulatory reports on time
We went from waiting two days for reports to having live dashboards. It fundamentally changed how our leadership team makes decisions.