Social platform development (Ezebee) | CASE STUDY

Ezebee — V02

For the first Ezebee version, we used an MVC architecture based on Zend framework, with a relational SQL database.

To achieve the client’s improvements needed for version 2, we adopted a service-oriented, RESTful, Web API architecture based on a Sails.js framework, in Node.js and the back-end relied on OrientDB (a NoSQL graph database) which enabled the use of Elastic Search. The UI was implemented in Knockout.js and we used Web Sockets to support real-time chat.

We used Braintree and PayPal APIs to allow money transfers, directly through the chat application.

We created a dynamic search system using tags’ popularity scores, preloading results, suggesting criteria to refine the search and loading the results at a very high speed. The graph database and the tag weighting system — depending on how many products are associated with that tag — enabled faster and more flexible categorisation of products.

Both v1 and v2 used MySQL, jQuery, CSS3, and HTML5, as well as Amazon Web Services hosting.

Originally published at https://www.roweb.ro.

Senior Marketing Specialist at Roweb 10+ years of experience in digital marketing, highly skilled in content writing and inbound marketing.