How Satin Creditcare Gained a Smarter Customer Lens

Satin Creditcare Network (SCNL), India’s third-largest Microfinance Institution (MFI) in terms of AUM and borrower base, realised in 2023 that it was sitting on a goldmine of customer data spread across loan systems, CRMs, and payment gateways—but couldn’t extract the contextual intelligence. The leadership wanted a tool that could help frontline teams act faster, personalise better, and reduce risks in real time.

Siloed data

Data was all over the place—behavioural, demographic, and transactional details lived in different systems. Moreover, it resided in different formats. The data availability was also not real time. It was batch processed, leaving various teams to strategise and take decisions based on old data.   

The first job was to bring them together. The concerned teams built a data lake to centralise raw inputs, designed Extract, Transform, and Load (ETL) pipelines to clean and unify them, and created unique customer IDs with master data logic. To make things smarter and quicker, they added APIs and real-time Kafka-based event streaming. With these pieces in place, Satin could finally get a 360-degree customer view, leading to sharper underwriting and smoother operations.

Once the data was ready, the next step was to make it work harder. The data science team developed segmentation models using unsupervised learning methods like K-Means and Hierarchical Clustering. They mixed behavioural cues with demographics and engineered clever features like transaction frequency and average ticket size. 

“But they didn’t stop at just training models, they validated the segments with domain experts, deployed them into business systems via APIs, and set up regular monitoring to keep them fresh and accurate,” says Sunil Yadav, CIO, Satin Creditcare.

Smarter Buckets, Smarter Business

This ML-led segmentation helped identify from high-value loyal customers to those at risk of default. It led to more relevant product recommendations, a sharper eye on delinquencies, and ultimately, better portfolio health. 

“The rejection rate dropped. Risk exposure reduced and the whole process became more proactive,” says Yadav.

So what changed at the functional level was decision making using real-time dashboards. Now, credit teams get instant insights for faster approvals. Collections can spot issues before they are red flagged and marketing has the data it needs for targeted campaigns. Everyone, from risk officers to business heads, is working off the same real-time truth. 

It’s important to note that as per the FY25 financials, Satin Creditcare had a total of 32,87,098 customers, of which 8,29,969 customers were added during 2024-25. The core products disbursement in this year (FY25)  was to the tune of 9,837 crores.  

Seeing the Impact

The impact was visible. “Customer retention improved by 18%. Delinquency risk dropped. Decisions that used to take days now happen in hours. Cross-sell success went up, thanks to smarter targeting,” informs Yadav. 

The company’s CMD, Dr HP Singh, also acknowledged the contribution of data driven decision making in his message to the shareholders in FY25. 

“It is in the realm of credit quality and risk management that Satin’s differentiated model truly stood apart. While many players contended with elevated delinquency levels, we delivered one of the most commendable performances across the industry. Our Portfolio at Risk (PAR 1) ratio improved significantly, from 6.8% in September 2024 to 4.9% by March 2025, well below the sector average. This improvement was not the result of retreating from specific geographies or restricting disbursals. It was achieved through disciplined, data driven borrower engagement, agile front-line response, and targeted, risk-aware field interventions,” he says. Moreover the PAR 90 ratio also improved to 3.3% until March 2025.

The platform, a web-based application went live in June 2024. It’s hosted on a public cloud. The team chose a microservices-based, cloud-native architecture. This approach allows the system to grow with the business and integrate with other tools down the line. From encryption to RBI-compliant data policies, every regulation box is ticked. It also provides for a centralised metadata, audit trails, and access control to ensure quality and accountability at every level.