In the context of updating the existing website and adding new functionality, our team took the following steps:
●
Designed and built new web pages, including “car review” pages, and “in stock lease deals” functionality, which allows clients to search for the right car based on budget requirements, fuel type, and other parameters.
●
Deployed micro front-end development architecture to transfer the “checkout” page and all the related functionality into a separate module. We designed the corresponding web pages and implemented the business logic behind them.
To optimize the vehicle leasing software performance, we accomplished the following:
●
Used Lighthouse, a web app audit tool, to perform tests and track performance, accessibility, and speed metrics of different web pages
●
Compressed some icons and other graphic elements to speed up loading
●
Migrated from Apollo Gateway to Apollo Router, which made queries on the website significantly faster as the Router’s response time is close to that of direct querie
●
Used lazy loading strategy to only load the parts of a web page that are being displayed on the screen, and load more components as the viewer scrolls through the page. This approach decreased the web pages’ loading time and made the displayed parts immediately available for interaction without waiting for the entire page to load.
●
Performed end-to-end (E2E)
automated tests for popular flows, such as search and filtering, to make sure that our optimization efforts didn’t compromise functionality. Our team designed and documented these tests, and integrated them into the CI/CD pipeline for all environments except the production. During the E2E tests, we also discovered and fixed some critical errors, such as mistakes in the leasing price calculation.
Our team also worked on document parsing. Originally, all the documents related to leasing and insurance were processed manually, and human errors were common. Our developers automated document handling and enhanced the process with Amazon Textract – an intelligent document processing service, which includes a machine learning component for document analysis.
