7 December, 2023
Cashdrive is an online loan service secured by a car. The company launched a new underwriting system - risk assessment before issuing a loan. Cashdrive's new underwriting takes into account the geography of vehicle operation and uses artificial intelligence to assess risk. Underwriting consists of dozens of types of parameters that are entered into the database automatically. The system takes into account fraud indicators and evaluates risk factors based on data from:
The AI module of the underwriting system consists of algorithms and neural networks that are specially trained on verified data sets. The new solution helps service employees evaluate the client’s profile according to the maximum possible criteria, highlight risk factors, calculate loan rates and evaluate the value of collateral. The AI module assesses risks based on geoanalytics. Using GIS Atlas blocks, the system calculates the risks when operating a pledged car depending on the city and area of residence. Geodata helps to correlate, for example, the city of driving with the quality of roads and the traffic police database: the types and severity of accidents in this area, the number of administrative violations and car thefts in the region.
“We work online almost throughout Russia. It is important for us to evaluate parameters for individual regions and cities with different socio-demographic characteristics. This is necessary to make a decision on issuing a loan and to calculate the optimal coverage of the loan amount with a collateral car. Of course, a similar process exists in all banks and microfinance organizations that engage in collateral lending. But we are glad to have additional expertise on our side, also fully automated. Our partner Marketing Logic's AI modules use dozens of layers of data from the GIS they developed. For us, this is a significant advantage in underwriting, which does not require us to develop all the blocks from scratch. At the same time, we receive the most complete predictive analytics, which takes into account dozens, if not hundreds of factors. This cannot be achieved using solely the expertise of our employees, so we have introduced modern technologies to help them,” notes the cashdrive company.
The underwriting and risk calculation system is seamlessly integrated into the CRM: the process is displayed there step by step with all the data. An employee directly in CRM sees documents, data, calculations and coefficients that the system processes.
“The cashdrive project turned out to be interesting and ambitious. We not only generated new geofeatures for predictive models, but also participated in the development and configuration of elements of the credit conveyor. Geoexpertise here enriches and refines the data, helps in each case to rely not on abstract or average indicators, but to apply them to a specific city or region. In the project, we applied our many years of experience in working with banks and credit risks, as well as technical expertise and the ability to work structurally with data. We are grateful to the cashdrive team for their trust and plan to develop our cooperation,” says Marketing Logic CEO Dmitry Galkin.