Balance Hero, Runner-up in International Financial Data Science Competition

Korean fintech company Balancehero is gaining attention by proving its technological prowess on the global stage. It secured second place in the 'African Credit Data Model Development Competition' hosted by Zindi, a network of data scientists from 195 countries. Competing against approximately 1,000 global data scientists, it was the only Korean company to win an award. The core task of this competition was to build an AI-based machine learning credit assessment model applicable to Kenya and Ghana, developing a universal system to predict the likelihood of loan defaults in developing countries. Balancehero successfully solved this challenge through its proprietary Alternative Credit Scoring (ACS) system. This machine learning-based system, which analyzes user smartphone data and accumulated user data, recommends and matches loan products optimized for users. Through this, it has shown innovative results, increasing loan approval rates by more than 10 times and reducing risks by up to 40% compared to traditional screening methods. Jeon Hyeon-woo, a data scientist at Balancehero, attached great significance to proving that the ACS methodology, validated in India, effectively works in entirely different regions like Africa. CEO Lee Chul-won emphasized that this award not only proves Balancehero's data analysis and AI technology are world-class but also confirmed the potential of ACS to expand beyond India to other developing countries. The company plans to continue its growth by providing better services in the Indian financial market through continuous technological innovation. Indeed, Balancehero has already proven its capabilities in the market by achieving its highest-ever revenue of 84.5 billion KRW and operating profit of 16 billion KRW in India in 2023, as the first Korean operator to do so. This international award demonstrates that Balancehero's technology has laid a crucial foundation for providing seamless financial services across borders.