Rapid data analytics and reporting platform
Our client provides transportation spend management solutions for various industries worldwide. These solutions include freight invoice audit, contract optimization,and business analytics.
The legacy reporting platform used by our client was a web portal that enabled them to schedule and run reports. This platform was backed by Crystal Reports technology running against an Oracle database that contained shipment invoice data.
In recent years, a steady increase in the volume of shipment invoice data feeding into our client’s systems overwhelmed the legacy reporting platform. Backed by a traditional RDBMS tech stack, it was unable to scale effectively for analytics and reporting. The performance issues resulted in delayed availability of reports, reducing the ability of the end-users to gain actionable insights into their data and take business decisions.
The key goal of this project was to dramatically improve the performance of the reporting platform. Executive-level reports that summarized millions of rows of data needed to be generated within seconds.
Filtering capability that included an “as-you-type” search was required for users to slice and dice their data quickly.
Another important goal was to provide a smooth user experience, with intuitive navigation and a focus on self-service.
The new reporting platform, VERNA, was built with a tech stack that is more suited for rapid data analytics and reporting. Elasticsearch was chosen as the datastore, as its distributed architecture with fast search and aggregation capabilities were a good fit for this use case.Angular JS was the front-end technology and backend services and jobs were built with Java and Spring Boot.
Tools were developed to automate the migration of shipment invoice data from the Oracle database into Elasticsearch. This included historical data going as far back as 3 years and newer data that came in daily.
Awesome … scalable, distributed and truly elastic application that the operations team love !