Midland Credit Management Boosts Scalability 50x, Cuts Costs 120x with MongoDB and AWS

As part of its commitment to treating consumers with empathy and respect, specialty finance company Midland Credit Management, Inc., (MCM) sought to build an omnichannel solution to improve both the consumer and agent experience. Working with Amazon Web Services (AWS) and AWS Partner MongoDB, MCM migrated from a legacy system to a distributed, hybrid, serverless architecture capable of supporting several omnichannel use cases. Since modernizing its system, MCM has seen significant improvements, including a 50x increase in scalability and a 120x execution cost reduction for its marketing letters solution.

MCM, a subsidiary of Encore Capital Group, Inc., helps consumers resolve past-due debts. Because MCM specializes in accounts that have fallen behind and been charged off by the original lender, most of the people it serves are in a difficult and distressing financial situation. “Our mission is to create pathways to economic freedom for our consumers,” said Ali Montazer, chief technology officer at MCM. “We’re heavily focused on the consumer experience and making sure we treat people with empathy and respect.”

All points of contact, from marketing letters sent by MCM to inbound communications handled by MCM agents, must be timely, error free, and consumer oriented—not asking people to repeat sensitive information as they move from chat to phone, for example. MCM believes consumers should have a seamless, omnichannel experience customized to their specific situation. “We want to give them that and, at the same time, improve the experience for our account managers,” added Montazer.

The challenge was MCM’s on-premises IT infrastructure, which was built on batch and offline data extract, transform, and load (ETL) processes. “Our system couldn’t efficiently process all the data we gathered each day or share it across environments,” said Montazer. “We didn’t have the foundation we needed to deliver a high-quality consumer experience. To achieve that, we knew we had to modernize our architecture.”

To start its modernization journey, MCM turned to AWS Partner MongoDB and its developer data platform, MongoDB Atlas. Atlas provides a tightly integrated collection of data and application infrastructure building blocks that enable teams to quickly deploy bespoke architectures to address any application need. It supports transactional, search, and analytical workloads across mobile, distributed, and even serverless architectures, all while reducing data infrastructure sprawl and complexity. Through the AWS ISV Workload Migration Program, MongoDB and AWS have partnered to simplify the migration process and reduce costs for customers interested in migrating on-premises workloads to the cloud.

With the help of the MongoDB and AWS teams, MCM built a distributed, hybrid architecture that uses microservices and serverless. The new architecture makes it possible to create an accurate, unified view of the customer from siloed on-premises data sources. The AWS Database Migration Service (AWS DMS) change data capture (CDC) feature pulls data from the on-premises database as it changes, for example, when a customer makes a payment. That data is streamed using Amazon Kinesis and aggregated in Atlas. Next, Atlas processes the data and generates change events that are received by AWS Lambda serverless functions for each omnichannel microservice, such as email, SMS, letter, or call. In addition, Atlas provides data storage for the Lambda functions.

“This architecture enables us to push processing time and power to the cloud, where we can easily scale up and down as needed,” Montazer said. “At the same time, it improves the performance of our core on-premises applications. And, by aggregating data from multiple sources, it gives our agents an accurate view of the consumer that leads to a better experience.”

While modernizing its architecture was important, shifting the culture at MCM in support of this change was just as critical. There were processes built around the old architecture that needed to be changed or eliminated. For instance, people would write their own queries that locked the system, preventing agents from accessing the data they needed when talking to consumers.

Modernizing the culture required a multi-year approach that started with migrating a few simple use cases. Once MCM got comfortable with the new architecture, it engaged the MongoDB professional services and product teams as well as AWS to migrate increasingly more complex use cases. “We don’t look at MongoDB as a vendor—they’re a partner for us. Same with AWS. They helped us achieve the goals we set, and we appreciate that,” Montazer added.

So far, MCM has implemented serverless solutions for emails, marketing letters, and SMS. In the case of marketing letters, the results have been significant. While the legacy system could process only 14,000 letters in a single run, the serverless system can process 720,000 letters, a 50x improvement in scalability. Processing one million letters, an MCM benchmark, took five days using the legacy system; the serverless system can do it in 16 hours, or 5x faster. Further, 450 executions were needed per month to meet expected marketing volumes using the legacy system; today, it takes only four executions, which is 120x less expensive in processing.

“What’s more, the new solution with the right level of monitoring and error correction is flawless,” said Montazer. “We have 100 percent perfect runs. In a business like ours, where even a single error is cause for concern, this is a big achievement.”

The combined capabilities of AWS and MongoDB Atlas have also enabled use cases that were impossible with the legacy system. For example, calling consumers is still a large part of MCM’s business model. The company relies on external data sources for those phone numbers that may return as many as 10 listings per individual. Because the Consumer Financial Protection Bureau limits the number and frequency of debt-related calls, it’s critical for MCM to have the right contact information from the beginning. “We now have a system that optimizes the process,” Montazer shared. “We couldn’t do that efficiently before we moved to the cloud, and our business is benefiting.”

MCM expects to move even more of its legacy solutions to the new architecture, such as some of its internal legal management system. “In a short amount of time, we’ve built the foundation needed to implement more advanced use cases,” Montazer concluded. “The next time we want to add something to the mix, working with MongoDB and AWS will be at the top of our list.”