In a high stakes market with complex processes, Auto Approve® assists consumers in refinancing auto loans with an extensive lender network across the United States. Using microservices architecture on the Microsoft Azure, the “Approve Engine” application is architected to scale and optimize the overall business process and value stream. As a result, within the first 12 months of deployment, Auto Approve® more than tripled the number of loans closed per month without needing to substantially increase the number of staff.
Auto Approve® a nationwide auto loan refinance company founded in 2016 and based in Maple Grove, MN. They built a groundbreaking refinance platform “Approve Engine” entirely powered by Microsoft Azure to assist vehicle owners in reducing their interest rates and payments. Auto loans reached almost $1.2 trillion in 2019, up 6.5% over 2018, and makes up 9.5% of American consumer debt according to Experian.
“A key Auto Approve objective was to be able to keep up with the rapid market growth,”
Jerry Kroshus, CEO Auto Approve®
TechFabric strategically broke down the legacy monolithic application into a set of event-driven, reactive microservices. Each of these microservices are responsible for automating business processes and collectively drive the “Approve Engine” platform.
Microservices are loosely coupled and reactive by design. Many subscribe to events happening in the system, process them, and broadcast any state changes to Azure Service Bus for other interested microservices to subscribe. This approach dramatically improves Approve Engine’s ability to react to changes happening in the system and ultimately set the groundwork for wholesale automation of the business.
In order to support large volume of transactions going through the system and meet the performance requirements, CosmosDB was selected as the database. It’s a globally replicated database with many consistency levels to choose from. It’s used as a system of record for storing all the master records and transactions. Any changes made to the documents in CosmosDB are synced to Azure Search in real time using ChangeFeed feature of CosmosDB.
That way, all read requests are sent to Azure Search while all the writes are sent to CosmoDB, thereby maximizing throughput and taking advantage of read-optimized query engine like Azure Search. By implementing this CQRS pattern, the system is fast, robust, secure, and scalable.
Azure Search was selected for its excellent support for faceted navigation, search suggestions, synonyms, accented characters, and fine tuning of results through scoring profiles. In a Refi platform, relevant search results to agent’s queries is perhaps the single most important factor affecting loan closing rates. Better search results enable agents to close more loans faster by providing better information to customers more quickly.
Using a combination of these features, the new Search API vastly outperforms the previous search system. Azure Search yields the most relevant results to a query based on a variety of attributes, with pre-configured weights and boosted values in scoring profiles. By carefully assigning weights to specific attributes, the search engine is configured, and continually optimized, to return optimal results.
To keep customers engaged during the Refi process, Logic Apps provides deep integration with phone and email systems used by Agents. With Logic Apps’ connectors for Office 365 and Twilio, all emails and text messages sent or received by customers are instantly displayed on the agent dashboard via Azure SignalR service. This way, all customer communication is preserved and can be used later for follow up events. Overall, Logic Apps and Azure SignalR plays a critical role in closing more loans.
“Microsoft technology helped us triple our business in under a year.”
Jerry Kroshus, Auto Approve® CEO.
Get in touch with our sales team, let’s start the conversation.