Bardess Group and Tokyo Century USA
Microsoft partner Bardess Group, based in Randolph, New Jersey, offers a pre-built platform comprised of technologies that promote the democratization of data. The platform is designed to solve modern scale problems and rapidly deliver business results.
Tokyo Century USA provides leasing finance to vehicle dealerships in the United States. During underwriting, Tokyo Century USA needs to determine the creditworthiness of potential clients and assess the risk of the loan. In the past, most of this was done manually, involving spreadsheets and sources that might not be updated regularly. The process could take up to an hour, with the potential customer waiting for a response. Tokyo Century USA was aware that response time could impact which lender the customer chose.
The four key components of the Zero2Hero Analytics Stack on Azure
The Zero2Hero Analytics Stack utilizes Azure Virtual Network, Azure Network Security Group, and Azure Blob storage. Tokyo Century USA chose the Zero2Hero Analytics Stack because it provides four key components required for analytics: data storage and processing, data preparation, data visualization, and advanced analytics. For the first component, Tokyo Century USA selected Cloudera Enterprise Data Hub, provided as a fully managed service. This gave Tokyo Century USA a turnkey data lake.
For data preparation, data from multiple systems was ingested into Cloudera Hadoop. From there, an iterative process of feature engineering was done by selecting data that business users felt would be predictive of credit quality. For data visualization, Tokyo Century USA selected Qlik Sense. Qlik Sense provides an open architecture that supports multiple consumption modes, including interactive dashboards and embedded analytics. For advanced analytics, DataRobot provides Tokyo Century USA with an easy-to-use automated machine learning framework. DataRobot provides an unbiased, data-driven assessment of risk, facilitating the work of credit risk analysts.
Having a consistent and systematic approach lets the company leverage the knowledge of its most experienced credit risk analysts and implement that knowledge in the machine learning models to the benefit of the whole department.
The proof of concept was successfully completed in 60 days and accelerated risk assessments from one hour to 10 minutes. The successful proof of concept cemented Tokyo Century USA’s commitment to purchase the software and services for a production system with a positive return on investment. Tokyo Century USA now has a modern data platform that can be expanded with additional use cases. The emphasis on democratizing data and providing business users with easy access allows them to provide direct input into the design process and accelerate innovation at Tokyo Century USA.