Project Background

This app was for smart city. It integrated a bunch of mini-programs so that citizens can use kinds of mini-apps in one super app. It promised user’s privacy with one-time login by OAuth protocol.

Business Introductions

The app showed mini-programs and news that people are interested in. It also recommended content personally based on user’s histories.

Mini-programs were pre-setted with multiple tags.
Data platform calculated user’s behavior asynchronous.
App listed the mini-programs that were with similar tags of most used mini-programs.

Technical Implementation

The app was built with hybrid mode, using native languages(Swift and Kotlin) as JSSDK to build native capabilities like camera, photo albumn and so on. The UI was built with React.js, it ran in a webview which was provided by the native layer.
The backend was developed with micro-service mode on AWS. Backend provided RESTful APIs to frontend. Data Platform was mainly for personlized recommendatation, which is streaming computing implmented with Spark.