Kapital Bank: Estimating Customer Income With Behavioral Scoring
Kapital Bank: Estimating Customer Income With Behavioral Scoring
Kapital Bank built a behavioral scoring model in Dataiku to identify small- and medium-sized businesses for which to provide loans.
Share
3x
Faster than previous legacy models
Addressing a Market With Little to No Data History
Kapital Bank is on a mission to support small and medium-sized businesses and help individuals fulfill their dreams by providing loans and granting credits to those without regular incomes.
They operate on a solid financial foundation, committed to creating value for the future of their customers and employees. One of Kapital Bank’s goals is to attract top talent through an innovative approach and enhance brand recognition, positioning itself as an attractive employer.
The Solution? Building a Behavioral Scoring Model in Dataiku
Kapital Bank has implemented a cutting-edge behavioral scoring system to estimate customers’ financial capabilities based on their transactional data. The bank uses both customer transaction data and government-provided information to gain insights into customers’ income patterns.
By analyzing different types of transactions, such as daily shopping, luxury shops, and more, Kapital Bank can estimate a customer’s income accurately. To further mitigate risk, the bank cross-references this income estimation with its risk technology tribe, distinguishing customers with official salaries from those who might receive money from other sources like parents.
The calculated income estimation is then passed to another risk model that assesses the probability of default. Based on this model, customers receive a risk score that helps the bank evaluate the potential lending risk. Depending on the bank’s strategy and government regulations, the bank decides to provide credit to customers, with the maximum loan amount capped at 70% of their estimated income.
A Deep Dive Into the Model
The primary target audience of this scoring model includes individuals who typically receive cash payments and have irregular incomes, such as drivers, and cleaners. The average loan amounts typically range from 300 euros to 1,000 euros.
The project leverages the capabilities of the holding platform to gather transaction data from retail, luxury stores, and other sources. Throughout the implementation, data privacy remained a paramount concern. The bank ensures data security by hashing and anonymizing customer information, instilling confidence among users that their data is protected.
The behavioral scoring model incorporates different data sources: transactional data, data from the marketplace app of a loyalty system, and other data sources. Kapital Bank’s team of data scientists currently consists of 22 employees, while the overall holding employs around 65 individuals.
All About Tangible Efficiency
Compared to previous legacy models, the behavioral scoring system is approximately three times faster, streamlining loan approval processes. While the current focus was on being innovative and driving progress for the company’s growth, Kapital Bank may explore attaching specific KPIs to assess cost savings in the future.
Behavioral scoring propels Kapital Bank’s innovation by surpassing income-based scoring and uncovering hidden customer segments based on transactional behavior. This customer-centric approach extends credit opportunities to individuals who were previously ineligible, giving us a competitive edge in the market.
— Mehti Aslanov, CDO of Kapital Bank
The customer base growth, the number of credits issued, and the overall business value resulting from this system will be determined and reported by Kapital Bank in due course. Nonetheless, the implementation of behavioral scoring represents a significant leap forward in Kapital Bank’s talent acquisition efforts and showcases the organization’s commitment to innovation and customer-centric solutions.
BMO's AI-driven solution to analyze client calls is powered by Dataiku and has enhanced customer engagement and operational efficiency, earning global recognition.
Davivienda uses Dataiku to power data and AI projects across the business, from providing product recommendations for low-income clients to collections strategy optimization.
To protect investors and market integrity, FINRA uses Dataiku to swiftly analyze vast market data, preventing misconduct and fostering innovation while saving costs.
AI Governance and MLOps features in Dataiku allow machine learning engineers at a leading financial services firm to support more than 125+ stakeholders with mission-critical workloads at scale for detecting potential risk.
Bankers’ Bank leverages Dataiku to increase efficiency and ensure data quality across an array of financial analytics, ultimately reducing the time to prepare analyses and deploy insights by 87%.
