Alternative Credit Rating

Pulse

The “Pulse” is an alternative credit rating algorithm that can innovatively complement the traditional credit evaluation framework.
Even without financial information, the system can provide accurate and real-time assessments of the SME credit risk by collecting a myriad of open (public) data, generating RFM variables, and implementing multidimensional analysis.

Mechanism

RFM Processing
The “Pulse” system collects more than 300 data points for a single company and processes each data in terms of RFM (Recency, Frequency, and Magnitude) by analyzing the data generation log.
Significant Variable Selection
The system examines the correlation between the individual processed data and company default risk by executing an array of statistical tests (t-test, chi-square test, ANOVA, etc.) and defines the final variables according to the results.
Alternative Credit Rating
The algorithm scores individual companies based on the risk predictions derived from the RFM pattern. The system assigns each company to risk level categorizes and transmits information to financial institutions.
Key Cases
Contract
Shinhan Bank
2024.09 – 2025.01

Developed a loan demand prediction model for high-performing companies and received the Seoul Mayer’s Award.

Contract
KBIZ Korea Federation of SMEs
2024.07 – 2024.08

Developed a business closure prediction model using mutual aid and public data for small enterprises and microbusiness.

Contract
Korea Credit Information Services
2023.12 – 2024.07

Developed an AI-based TCB (Tech Credit Bureau) guidance model by integrating internal and external data.

Contract
Lotte Card
2023.05 – 2023.07

Conducted performance validation targeting corporate card borrowers and the Pulse system is now linked to the Lotte Card at a real-time basis.

Contract
Woori Card
2023.05 – 2023.07

Conducted performance validation targeting corporate card borrowers.

Contract
Industrial Bank of Korea
2022.01 – 2022.08

Completed the performance validation for SME loan borrowers through the 1st LAB PoC (Proof of Concept) and secured a data supply agreement.