Upbound Group builds its modernized point-of-sale platform on AWS

Upbound Group, Inc. (NASDAQ: UPBD) is an omnichannel-platform company headquartered in Plano, Texas. It’s committed to providing innovative, inclusive, and technology-driven financial solutions that address consumers’ evolving needs and aspirations. Upbound’s customer-facing operating units include industry-leading brands such as Rent-A-Center® and Acima®, which facilitate consumer transactions across various store-based and digital retail channels. Upbound has more than 2,400 company-branded retail units across the United States, Mexico, and Puerto Rico.

Adapting to changing needs

In the fast-paced retail world, technology is crucial to driving efficiency and customer satisfaction. Upbound realized that its legacy Swing-based SIMS point-of-sale (PoS) system was hindering innovation.

After considering the following key factors, Upbound decided to pursue an all-in cloud implementation:

• Outdated technology
  • Swing-based system used limiting, outdated technology
  • Difficult to scale and integrate with modern SaaS solutions
  • Requires workarounds, resulting in prolonged development and testing cycles
  • Creates technical debt and increased friction among coworkers
• Incoherent customer information
  • Lacked a unified view of customer profiles and interactions between physical stores and ecommerce business
  • Often led to unsatisfactory customer experiences and increased risk for fraud
• Operational confusion
  • Lacked end-to-end operational visibility for the legacy PoS system
  • Slowed down issue diagnosis and resolution

The following diagram outlines the digital-transformation framework Upbound used. This tech stack comprises a modernized API and an event-driven microservices architecture. It provides the scalability and elasticity needed to support Upbound’s growing customer demand.

Adopting event-driven, serverless architectures

Prior to modernization, Upbound used a monolithic architecture with on-premises legacy infrastructure. This prevented scaling per changing customer demand. It also stopped developers from building rapid features as needed. Upbound transitioned to a microservices architecture by implementing an API-first design using Amazon API Gateway, AWS Lambda, and AWS Fargate. It completely modernized its rent-to-own and enterprise-services app stack. Development teams can now quickly release new features and functionalities to enhance the shopping experience. And business teams gain the freedom to mix and match products to meet strategic and tactical business goals.

Upbound had 45 TB of historical data, with more than 2,600 tables, including 1,080+ transactional tables, legacy lease agreements, and archival data. It wanted to break free from commercial licensing, stop self-managing database servers, remove data silos, maximize data, and share data across applications and organizations. Upbound used AWS Data Migration Service (DMS), AWS Schema Conversion Tool (AWS SCT), and custom database development to migrate from its legacy transactional database (running on Oracle Exadata) to purpose-built Amazon Aurora PostgreSQL.

Then Upbound optimized, redesigned, and consolidated its transactional data model by prioritizing normalization and efficient indexing. This enhanced application performance, reducing the number of tables to 400 and the data size to less than 30 TB. Upbound also upgraded its Aurora instances to AWS Graviton3-based R7g database instances, improving overall system performance. Now data grows in a way that’s steady, predictable, and cost efficient.

Selecting flexible, scalable integrations

Upbound chose Amazon RDS Proxy, a fully managed, highly available database proxy, to integrate its core business and enterprise microservices with RACDB, the primary transactional database. RDS Proxy connection pools helped decouple and scale Lambda microservices and Aurora databases separately. This eliminated resource and memory contentions from oversubscribed connections, because Amazon Simple Notification Service (SNS) topics fan out multiple Amazon Simple Queue Service (SQS) queues for asynchronous processing needs.

To improve the omnichannel customer experience, Upbound added a Global Customer Database (GCDB) using Amazon DynamoDB and search capabilities using Amazon OpenSearch Service. This applied to its customer-facing applications, too, including the modernized PoS RACPad. New customer transactions in RACPad are also updated in GCDB, ensuring a single data source for customer touchpoints. A single view of customer data lets Upbound engage with customers using personalized marketing campaigns, integrated customer support, cross-selling, and upselling.

Enhancing end-to-end security, availability, and resiliency

To improve availability and reduce downtime, Upbound deployed its tech stack in multiple AWS Regions in a Pilot Light configuration. For the data layer, Upbound adopted Amazon Aurora Multi-AZ clusters with Read Replicas. For the storage layer, Upbound uses AWS S3 Cross-Region Replication to replicate its business and application data from a primary to secondary region. DynamoDB Global tables are used to manage cross-region table replication, and change-data capture for DynamoDB streams keeps the primary and secondary GCDB in sync. CI/CD pipelines are configured to deploy application packages to the secondary location and integrate them with secondary database clusters. And platform engineering teams use AWS CloudFormation stacks for infrastructure-as-code deployments.

Upbound also relies on several other AWS services, such as the following.

• AWS Control Tower creates, manages, and scales AWS landing zones.
• AWS Direct Connect facilitates dedicated, hybrid network connectivity between its private infrastructure and AWS.
• Application developers use Amazon CloudWatch and AWS X-Ray to enable custom metrics, custom logs, and tracing to troubleshoot and monitor application performance.
• Database admins use Performance Insights for advanced performance monitoring and to tune their Aurora PostgreSQL clusters.

Reviewing outcomes

Upbound creates and manages AWS CloudFormation stacks for templated infrastructure provisioning. AWS Config maintains AWS resource inventory, configuration history, and change notifications for enhanced security and governance. And platform engineering, application development, and enterprise services engineering teams collaborate effectively by forming multiple “Two-Pizza Teams.”

This transformation helped Upbound provide a consistent omnichannel experience for its customers. Unified customer profiles from Global Customer Database powered by Amazon OpenSearch help business teams create targeted, personalized campaigns for customer acquisition and retention. Upbound successfully migrated its extensive historical data, enhancing scalability, reliability, and operational agility with RACPad across 2,400+ stores and partner locations. This cloud-first architecture has streamlined processes and saved hundreds of hours per month across stores, freeing staff to focus on strategic tasks and improving customer service.

In our upcoming blogs, we’ll dive deeper into the technical components and design frameworks of RACPad.

“Building RACPad on AWS’s suite of services with a serverless-first approach offers a transformative experience. This shift allows for rapid development cycles, as new features can be deployed quickly without the constraints of infrastructure provisioning. One of the key advantages of serverless architecture is its ability to scale down to zero cost during off-peak hours. This is achieved through the pay-as-you-go pricing model.” —Pranav Sharma, Director of Platform Delivery

“When going from a traditional on-premises to serverless architecture, you [need to consider] the company’s technical capacity for the migration…. Leaning on the serverless community and AWS Enterprise Support helped [us] tremendously.” —Mike Porras, Director of Platform Engineering

Download Transforming Retail in the Cloud: A CIO’s Handbook to learn how AWS can help you grow your business. AWS is also a member of the MACH Alliance. It facilitates innovation and better supports enterprises, retailers, and CPG companies in delivering the best customer experiences. Learn more about AWS Retail Solutions here.