Clientโ€™s quote
โ€œThe implementation of AI-driven AML into our fraud prevention module decreased our false-positive transactions by ~80%.โ€

Introduction

First Bridge has created a proprietary neural network application for a fintech company to process data. The company's platform facilitates financial interactions between individuals and legal entities through a payment system.

Challenge

Around 15% of clients have been making late payments on their loans, and there was no system in place to assess risk and forecast payment patterns. This has led to delays in the decision-making process, and a significant human factor in analyzing client data, potentially resulting in the approval of large loans to companies with poor payment histories.

The project aimed to optimize existing processes, mitigate financial risks, enhance overall security, and expedite decision-making by incorporating AI. Additionally, the project aimed to minimize the impact of human error on the decision-making process.

Upper level plan

Together with the client, First Bridge R&D team has analyzed the biggest challenges, performed a research of the possible solutions and proposed to implement 3 process automation modules:

Anti-fraud
Risk assessment and forecasting
Analysis of credit history

Fuzzy systems technology was chosen for the first two modules, and neural networks were chosen for the credit history analysis module.