ID Finance has enhanced its credit scoring capabilities with the integration of its newly developed AI-based fraud-scoring engine. By eliminating fraudulent loan applicants, ID Finance has been able to improve the quality, precision and speed of its credit scoring technology, boosting its loan approval rate by 50 per cent.
“By retrospectively analysing fraudulent loan applicants and identifying various red flags, we’ve been able to develop a powerful algorithm which eliminates these cases early on. As well as helping us reduce fraud, it has allowed us to dramatically improve our loan approval rate without impacting risk,” comments data scientist Irina Khoroshko, who led the development of the fraud-scoring technology at ID Finance. “Ultimately it means we are able to increase financial accessibility by extending credit to more people and helping them rebuild their credit history.”
The fraud-scoring engine works by analysing information relating to the applicant’s authenticity and detecting atypical behaviour or anomalies in the data provided. The engine takes into account a wide range of factors including the time spent filling in the form as well as the applicant’s web browser history. The quality of supporting documentation uploaded is also assessed for authenticity.
By screening out potential fraudsters early on and revising the ‘stop factors’ used within the credit scoring process, ID Finance has significantly improved its loan approval rate, as well as reducing the cost of issuing new loans and the incidence of non-performing loans.
The fraud scoring system was developed in-house by ID Finance’s team of data scientists and risk analysts and based on meticulous analysis of previous fraudulent applications. The engine uses machine learning to examine over 15 external data sources and several hundred data points to identify patterns that would otherwise go undetected.
Boris Batin, co-founder and CEO of ID Finance comments: “This has been possible thanks to extensive collaboration between our risk and security teams as well as constant experimentation and reiteration of our data science tools. Our aim is to have the best performing risk management system in the industry and we are making excellent progress.”
The fraud scoring engine complements ID Finance’s fully automated scoring technology, which rapidly identifies prospective borrowers with low or no credit history but who demonstrate behaviours similar to customers with positive repayment histories.