This payments and payroll technology company had been working for over a year rewriting its integration APIs to facilitate a more cloud-native approach to their business that would allow different divisions within the company, as well as third-party vendors, to integrate external applications with the payroll system. Their monolithic APIs were difficult to maintain and very fragile ― any time a change was made, it would often create bugs elsewhere in the system because too many systems were entangled.
The company wanted to decouple the integration APIs and build a dedicated API using microservices. The purpose was to create a more agile, modern and cloud-native system-to-system data integration process to increase performance and scalability.
“They had a minimum viable product they were trying to finish – they wanted it out the door,” a Blueprint developer said. “In order to do that, though, they had to deploy the infrastructure behind these new microservices that were previously created manually. That’s where Blueprint came in.”
The Blueprint Way
Blueprint brought in dedicated development resources to join the company’s internal development team and get the project across the finish line. With the additional talent and bandwidth, the new team was able to build the necessary infrastructure quickly and efficiently as code to modernize the payroll API. When a company defines infrastructure as code, it removes the time-consuming manual process of creating the infrastructure required to run an application or service and then document those steps for future changes or reproductions.
“Infrastructure as code defines what is needed in code that can then be applied to automate the creation of resources in Azure,” a Blueprint developer said. “There is a clear-cut pattern to make changes, reducing the cycle time and the risk that something will be done improperly.”
Once the infrastructure was complete, Blueprint was able to help develop and deploy the microservices that would allow the company to improve the performance and scalability of its payroll API. At the same time the risk of bugs entering the system when changes are made was minimized because any change can now be made in isolation.
For this company, this project was also a test to establish what might be possible moving forward. Now that it has successfully decoupled the payroll integration APIs from its monolithic system and used infrastructure as code to build in the microservices, the company can continue modernizing its monolithic APIs, using this now-established pattern.