Main content Main content

Automated Credit Approval using AI Technology

Precision machine-learning automates the loan credit approval process.

Background

Eyes are on fintech – combinations of finance and ICT are producing innovation in financial services and can help refine and streamline conventional processes. Fintech is in demand particularly in the finance industry, where declining profitability due to lower interest rates has become a major concern of senior management.

Solution

In the past, credit evaluation was typically done manually. Ricoh has now developed an AI technology that automates the approvals process, and which reduces the workload of those involved in credit approval. Credit approval tasks, include determining the appropriateness and maximum amounts of loans can now be automated and better prediction of potentially unrecoverable loans leads to a reduction in losses.

Credit approval process takes time if done manually.

AI-based credit approval accelerates businesses and identifies probability of defaults.

Technical highlights

The AI predicts the probability of default and determines the appropriateness of extending credit by entering information (characteristic quantities) of the company you are analyzing, such as sales and business sector.

Ensemble learning is a type of machine learning where multiple learning algorithms are integrated as one. Ricoh's technology uses ensemble learning to precisely predict default probabilities and determine credit appropriateness.

该技术具有以下额外features:

  • Trends and patterns in characteristic quantities are automatically extracted and determined from the many possible combinations.
  • Prediction modelling ensures that the solution is robust, especially given fluctuations in the business climate, by using previous fluctuation records.

Concept of ensemble learning

Ricoh's vision

The Ricoh Group has many medium- and small-sized companies as its clients. AI technology is ideally suited for credit approval tasks for these companies because of its high degree of accuracy, which increases with the volume of data processed. Ricoh continues to use this technology in its own internal processes across its group companies and is planning to offer this service to other companies that have issues with their credit approvals processes.

*
This technology is used today by Ricoh Leasing Company, Ltd. in Japan.