Skip to main content

How Malaysia’s leading telco is exploring privacy-safe data collaboration for credit intelligence 

Results
0 M
previously unscorable consumers now assessable
0 %
improvement in credit decisioning accuracy for thin-file applicants
0 %
faster time-to-decision for BFSI partners via API
The Challenge

Extending credit access to unbanked and underbanked consumers required going beyond traditional bureau data. Many thin-file applicants lacked sufficient financial history to be accurately assessed using conventional credit scoring models.

To address this gap, telco network data presented a powerful alternative signal—offering behavioural insights such as payment consistency, usage patterns, and tenure. However, combining telco data with bureau scores introduced significant challenges:

  • Ensuring data privacy and regulatory compliance

  • Enabling collaboration between telco and financial institutions without exposing raw customer data

  • Building a scalable model that could integrate alternative data into existing credit workflows

  • Delivering faster decisions without compromising risk controls

The client needed a secure, privacy-safe data collaboration framework that could unlock new credit insights while maintaining strict governance standards.

The Solution

ADA designed a phased transformation roadmap—starting from secure data collaboration infrastructure through to production-ready credit scoring models.

The journey began with the implementation of a Snowflake Data Clean Room, enabling telco and BFSI partners to collaborate on anonymised, encrypted datasets without directly sharing personally identifiable information (PII). This privacy-by-design environment ensured regulatory compliance while unlocking high-value alternative data signals.

Next, ADA engineered enriched feature sets by combining bureau scores with telco behavioural attributes, enhancing risk visibility for thin-file applicants. Advanced analytics and machine learning models were developed to improve predictive power, with rigorous validation and bias testing to ensure fairness and accuracy.

Finally, the enhanced credit scoring model was integrated via API into BFSI partner systems, reducing processing times by 40% and enabling near real-time decisioning.

This structured, privacy-safe approach enabled financial institutions to responsibly expand credit access while improving underwriting precision and operational efficiency.

Results
Challenge
Solution
Get in touch

Let’s turn your data and AI into measurable outcomes