Dailypay, a financial solution provider for small merchants, has built a big data repository that collects and manages business data of small merchants and self-employed people in real time. It plans to use the accumulated data for credit scoring and calculating credit limits.

Dailypay said on Aug. 19 it has built R-SDB (Real-Time Small-business & self-ownership DataBase), a real-time big data repository that integrates and manages business data of online and offline small merchants and self-employed people, along with POS and VAN data.

To build R-SDB, Dailypay linked APIs of financial institutions and platforms. It combined OCR technology and LLMs (large language models) to structure documents so the collected data can be analysed, and it also applied data normalization logic to manage data from different sources under the same standards.

The company said this has shortened the time required for analysis and screening while processing more than 100 million data records.

Data stored in R-SDB is used for credit scoring, assessing risk and calculating credit limits by linking it with business owner information. It collects and stores business data in real time to track sales flows and changes in business conditions for small merchants and self-employed people.

Dailypay plans to use R-SDB to analyse sales transaction flows and volatility, business conditions and other factors for online and offline businesses and apply the results to its liquidity solutions.

It is also pushing to develop R-SCB, a real-time credit scoring algorithm specialized for small merchants and self-employed people. It plans to use the accumulated data to advance its credit scoring and credit limit calculation systems.

Dailypay CEO Hae-woo Lee (이해우) said the difference from small merchant SDB and SCB systems that have recently been discussed for adoption in the financial sector is that it collects, manages and evaluates real-time data. "We will collect scattered data and use it for credit scoring for small merchants and self-employed people," he said.