Part 2: CredoDecide – Integrated Digital Originations Channel
December 3, 2020
CredoLab's Chief Data Scientist on How Banks and Consumer Lenders Can Leverage Data Analytics for Business and Social Impact
In part one of the blog series, “Digital Originations: The Imperative Solution to Banking in 2020 & Beyond” we identified the bottlenecks that hinder the fast-track digital transformation of lenders and confirmed the existence of a high priority requirement for a pre-configured and pre-integrated solution that provides a rapid and effective digital channel for organisations.
In part two, we dive right into the solution – CredoDecide.
CredoDecide - An Integrated Solution
CredoLab and ADEPTTM Decisions have developed a fully integrated solution which will accelerate lenders digital transformation using a low risk, but rapid integration. CredoDecide combines the power of an intuitive and light-to-download mobile app, alternative mobile data, and a decision engine to enable customers to apply for a loan without having to visit a physical branch.
Let’s take a closer look at each of these elements.
CredoApply is a digital onboarding mobile app packed with device intelligence. A fully customisable and white-labelled app, CredoApply offers the ability to collect application form data, reduce KYC manual checks, perform anti-fraud checks, and assess applicants’ creditworthiness via CredoLab’s alternative data credit score.
CredoApply turns even New-To-Credit (NTC) or unbanked users without credit history from “thin files” into “thick files” by processing two types of data:
- Personally Identifiable Information (PII): This data is filled in the application form by the applicant and it is sent from the app directly to the client via a dedicated API. This data is used to perform host-to-host, real-time checks without being stored on CredoLab`s servers.
- Digital footprint: This data is anonymous and is processed by CredoLab in the app. The result is a JSON file that contains only metadata, i.e. data about other data. The JSON file is used only to calculate the credit score of a user and to find the correlation between them as an applicant and the predicted default rate.
CredoLab uses smartphone device data to generate an application score for the customer which predicts the probability of defaulting. By accessing first party, privacy-consented, permissioned and anonymised smartphone metadata, CredoLab’s AI-powered algorithm sends the financial institution the score of the applicant in a matter of seconds. In doing so, not only do we open a whole new segment of customers for banks and lenders, but we also bring down their cost of risk, speed up the processing time, and increase the overall Gini Coefficient.
Core to CredoLab’s algorithm is its proprietary data modelling pipeline. The data modelling pipeline consists of a series of automated steps, rooted in Machine Learning techniques, which ultimately outputs a predictive model for credit default. To protect the customers’ confidentiality and to prevent bias towards individual loan customers, CredoLab uses only non-personally identifying metadata. The algorithm considers over 50,000 data points before arriving at the score that predicts the creditworthiness of applicants. To address the different risks of customers in different markets and the different nature of industries and products, our data scientists work with each client to create an algorithm that best suits their particular business.
The fully integrated decision engine enables lenders to:
- Rapidly implement automated decisioning strategies through an easy to use interface.
- Conduct champion/challenger testing, which ensures a ‘test and learn’ dynamic business environment
- Determine acceptable risk levels to minimise unprofitable accounts
- Automate and test policy rules
- Implement additional application scorecards that utilise geo-demographic information.
- Determine product terms and conditions
- Implement risk-based pricing strategies
- Run simulations prior to taking a strategy live
- Dashboards provide a real-time view of the business and can be accessed 24/7/365.
The decision engine is hosted in the AWS cloud, and can be accessed by clients anywhere in the world and through any device. Changes in strategies can be simulated and easily deployed into a test environment or directly into production, without the need for IT resources to be involved. This solution puts the credit risk manager in the driving seat and enables lenders to monitor and react to change in a rapid and effective manner.
How does it all come together?
In our last part, we show you how these different elements come together to provide a powerful credit decision tool for the COVID-19 world. Stay tuned to complete the journey of digital originations with CredoLab and ADEPTTM Decisions.