[AI Revolution](53) AI Technology Assisting 'Jungsinyongja' Loans... PeopleFund

Establishment of AI Research Institute and Development of New Financial Technology
Resolving the 'Interest Rate Cliff' with AI Technology

In South Korea, it is a reality that individuals with medium credit scores find it difficult to obtain loans at reasonable interest rates. The threshold for banks with low interest rates is high, and turning to savings banks or loan companies significantly increases the interest burden. PeopleFund is a company that aims to solve this 'interest rate gap' problem using artificial intelligence (AI) technology. The reason non-bank sectors have high interest rates is due to risk costs, and the core is to enhance the technology managing these risks with AI to enable accurate credit evaluation and default prediction. The explanation is that if a high-quality group can be accurately identified among those lumped together as medium to low credit, the company can manage its risk while offering better terms.

Kim Dae-yoon, CEO of PeopleFund, said, "We have accumulated AI-based risk management capabilities," adding, "We established an AI research center in 2019, and over 60 engineers from prestigious domestic and international universities are dedicated to developing mobile-specialized credit finance technology." So far, PeopleFund has spent over 30 billion KRW on research and development. Just last year, about 8 billion KRW was invested to commercialize a credit evaluation model specialized for mobile finance.

Daeyoon Kim, CEO of PeopleFund

AI Technology Resolving the 'Interest Rate Gap'

PeopleFund's AI credit evaluation model is characterized by integrating AI technology even in the data preprocessing stage to enhance learning performance. Credit information is highly imbalanced between good and bad credit, and each borrower's default pattern is different. This peculiarity of credit information is critical for AI learning with high sensitivity. From the AI learning perspective, PeopleFund explained that they transform structures to increase the utilization of each variable and have built various automated data preprocessing solutions to perform data preprocessing faster and more precisely.

AI technology has also been applied to the 'fraud detection system.' This is because the non-face-to-face loan process has expanded vulnerabilities to fraudulent activities such as identity theft and loan stacking. PeopleFund operates algorithms that determine whether the ID submitted through the platform matches the image of the customer currently using the mobile phone, as well as algorithms that identify various fraudulent patterns through usage behavior within the platform. CEO Kim said, "PeopleFund's loan fraud prediction models, developed specifically for various types of mobile loan fraud, show higher performance compared to other financial institutions, along with a real-time credit strategy derivation system."

Additionally, PeopleFund developed and applied an 'explainable AI model' to the outputs of the AI credit evaluation model, which can provide objective and transparent grounds and reasons in a multidimensional manner. They also built an 'AI loan approval strategy automation system' that automatically derives the optimal loan approval strategy whenever the market changes. This system showed a 23% increase in loan approval rates and a 25% reduction in default rates compared to existing loan approval strategies.

More Than Half Developers Despite Being an Online Investment Finance Business

The foundation of this technological capability is excellent personnel. PeopleFund, while operating as an online investment finance business (Ontu-eop), has structured its workforce centered on engineers. Experts in data and product development make up 53.9% of the total. PeopleFund's tech team also won first place last year in a machine learning competition for the financial sector hosted by Amazon Web Services (AWS), outperforming major internet banks, financial institutions, and big tech companies.

PeopleFund believes AI technology can contribute to resolving the chronic 'interest rate gap' problem. A PeopleFund representative said, "AI technology will bring fundamental changes to the way financial institutions and credit rating agencies have managed risk so far," adding, "The more actively PeopleFund's solutions are used in the market, the more the overall financial market's risk management costs will decrease, and the saved costs can be used as additional resources to provide more reasonable financial products to consumers."