Case Study: Webbankir boosts automated loan decisions with KNIME

How Webbankir automated loan application decision making using KNIME

Webbankir, an instant online microloan provider in Russia, needed a faster and more reliable way to make loan decisions. Its existing process was slow to move models from development into production, and it required significant developer support to build and validate machine learning models. Webbankir turned to KNIME Analytics Platform and KNIME Server to create an automated decisioning tool that could work with external data sources, Python, R, and visual workflows.

KNIME implemented a cloud-based decision service on AWS, integrated with Webbankir’s existing systems, and enabled model development, validation, and deployment through PMML and automated testing. The result was a dramatic reduction in model implementation time from one to two months down to one to seven days, a 30% decrease in decision-making time, and an increase in fully automated decisions from 70% to 85%. The new service now handles more than 95% of the application flow and freed up one full-time software developer.

Webbankir

Sergey Bazov

Portfolio Manager