Build Small, Think Big -- Our Next Gen Platform
April 17, 2017
The world of financial technology is fast paced and unforgiving. Without the proper tools and know-how, it's easy for any company to get left behind. At Geezeo, we're planning for tomorrow with tools and strategies that have been proven in all areas of software engineering. Geezeo is building small and thinking BIG, utilizing much of the same technology powering Twitter, LinkedIn, Amazon, and many more.
Ten years ago, Geezeo began as a small application built on Rails, a technology platform that helps ensure rapid development. We also developed long-term relationships with MySQL as a datastore, and Amazon Web Services for hosting. As we grew (currently supporting more than six million end users) our team recognized the challenges associated with this monolithic architecture.
Operating at Large Scale
The first of the technical constraints was the stress the monolith was inflicting upon the MySQL database as the user base grew. Next, while Ruby is a really expressive language to write code in, large scale data processing is not a key feature of the platform. When looking to the future, the Geezeo team focused on ensuring the platforms ability to operate at a large scale.
Geezeo is now leveraging Scala, which runs on the JVM, to provide enhanced performance and flexibility to our new distributed platform. Under this new platform architecture, we’ve split each service, such as reading & writing transactions, into separate applications. This essentially breaks apart our monolith and reduces bloat and complexity while enabling us to scale these portions of the platform based on their unique performance characteristics.
These new applications, named for those of us who love food, include Pantry, StoveTop and Lazy Susan. Pantry is the stand-alone persistence application that allows us to streamline the process of pushing and pulling data from the platform. StoveTop is the workforce behind Pantry. It is a durable & flexible application for large scale data processing for our transaction pipeline and will be the backbone of our engagement, analytics and big data platforms. Lastly, Lazy Susan is Geezeo’s public facing API for data enrichment. This application is an additional option in our efforts to process all data from financial institutions to provide enriched transaction or account data along with improved data analytics in real-time.
Change Brings Consistency
Overall, these changes not only increase Geezeo’s performance but enhance our consistency, durability and scalability to ensure the end user receives the best possible experience while allowing our clients to leverage their customer data and explore the insights big data provides financial institutions.