Home / Case Studies / Indonesian Fintech DWH

Case Study · Fintech / Digital Lending · Cloud Data Warehouse

From reactive to real-time,
a lending business that can trust its data.

Client Confidential Indonesian Fintech
Sector Fintech / Digital Lending
Platform Google BigQuery
Duration 8 months

The challenge

The client was processing over 100 loan approvals every day, a volume that demands fast, accurate, and data-driven decision-making. But the data infrastructure underneath the business was not keeping up.

Transactional, KYC, payment, and loan origination data sat across seven separate source systems with no integration between them. There was no data warehouse, no unified view, and no analytics layer. Reporting was a manual exercise that produced results the following day, by which point the data was already stale. Fraud detection was entirely reactive: the team would identify fraudulent activity after the fact, with no systematic way to catch patterns before losses occurred.

For a lending business operating at this pace, the gap between when fraud happens and when it is detected is where money walks out the door.

What PGI delivered

PGI-Data designed and built a cloud-native data warehouse on Google BigQuery from the ground up, structured across three layers:

Between 300GB and 700GB of data was consolidated into a single, clean, queryable warehouse. A fraud detection layer was built on top using pattern recognition and rule-based logic across transaction history, shifting the team from reactive investigation to proactive intervention on active loan applications.

Measurable outcomes

7
Source systems
unified into one DWH
8-15%
Reduction in
fraud losses
Next day
→ Hours
Reporting
turnaround

Technology & capabilities

Google BigQuery Cloud Data Warehouse ETL Pipelines Data Modelling Fraud Detection Engine Pattern Recognition Analytics Dashboards KYC & AML Integration

Working on a similar challenge?

Talk to our Fintech and data specialists about your analytics or fraud detection programme.