Real-Time Decisions for Real Results
Enova Decisions real-time predictive analytics services help businesses improve the customer experience while protecting against fraud, optimizing operations and increasing marketing profitability.
By Joe DeCosmo, CAO of Enova International
As the Chief Analytics Officer at Enova, I’ve had a front-row seat to the power of predictive analytics in the online lending space. Predictive analytics are embedded in all that we do, and for more than a decade we’ve been predicting outcomes and automating decisions in real time to serve more than 4 million customers.
Seeing the difference real-time decisions can make in helping to drive meaningful results in everything from marketing to fraud reduction, our team recognized there was an opportunity to bring our analytics expertise and proprietary technology to the market. Through our real-time decision management service, Enova Decisions®, we offer on-demand decision-making technology and real-time predictive analytics services to businesses across industries including financial services, higher education and telecommunications.
A fast and effective way to make automated business decisions, Enova Decisions uses predictive analytics technologies to improve business functions and the customer experience by providing solutions for:
- Customer acquisition and retention through smarter product offers and credit risk decisions;
- Fraud protection and alerts through customer data verification, customized velocity scores and link analysis; and
- Operational improvements with optimized collections and payments practices.
The real-time analytics engine or platform that enables these solutions is called Colossus™ and was built using a best-in-class scientific computing platform for customizable scoring and decisioning in real time. Colossus also allows for the delivery of real-time analytics at scale, so that businesses can quickly scale up or down based on consumer demand.
Figure A: Colossus Real-Time Analytics Engine
Colossus can receive a model or decision request from any internal or third-party application, and return scores or results to the business within sub seconds. This happens through simple, web-based API’s that use JSON formats for input and output.
Figure B: Colossus Performance Snapshot
The system also is integrated with over 20 internal and third-party systems and data providers to quickly fetch, parse and calculate structured and unstructured data needed for model scoring and decisioning. Integration with multiple data providers or vendors allows for more precise predictive decisions. Imagine using only one data vendor— you have much less robust data on which to make a smart decision, as one data set only offers one view of a consumer.
Colossus runs a variety of algorithms from regression to machine learning, as well as simple decision rules, and it allows for rapid model implementation and improvement. Colossus deploys models built in SAS®, R, Python™ and other analytics platforms and environments. Because Colossus provides support for a wider variety of models, data sources and variables, it enables analysts to execute scoring code from multiple statistical platforms and greatly improves speed from modeling to deployment. It also gives analysts the ability to deploy models without software engineering, freeing up time and resources for all.
Enova Decisions is designed to provide business solutions for better answers in real time, and it has been tested and proven by Enova’s brands worldwide. As improving the customer experience has become top of mind for digital companies, real-time decisions that provide real results will continue to be important.