Case study
Tourie
Digital design, Development
Task
Design and implement an inteligent mobile city guide featuring elements of machine learning
Challenge
The goal was to create a mobile application (preferably available on various platforms as well) that would help with city sightseeing. Application should be able to generate a route according to user preferences, assist user along the way and learn user's habits.
Outcome
I created a simple and clean user interface, focusing heavily on ease of use. With small team of developers we decided on implementing the solution as a Progressive Web Application, which resulted in beautiful, efficient and platform-agnostic software. We delivered core features and expanded the power of the application with natural language processing and decision support.
Design
Let me guide you
From the start we knew that the application was predominantly geared towards mobile usage, but we didn't want to leave desktop users with nothing. That explains the mobile first approach I took when designing the application. The goal here was to make the experience clear and simple, yet friendly for the mobile user and be appropriately scaled for larger devices.
Progressive web application
Covering most platforms with one system
Want an application that works coherently on multiple platforms, from Android and iOS all the way to web regardless of the browser? Progressive Web Application is your best bet. It allowed us to make an application which is virtually indistinguishable from native mobile applications, works on any desktop device as a regular web app and finally is highly scalable, lightweight, accesible and efficient.
Development
Combining the newest with the most reliable
We used frameworks and technologies that are most widely used. Front-end application is powered with React, uses Redux for state management support and Styled Components for visual layer implementation. Back-end is a Node.js application organized in layers featuring a REST architecture. Finally, both applications are being continuously built, tested and deployed. It allowed for a reliable, scalable and easy to work with solution, that doesn't require any extra maintenance.
Machine learning
Adding a sparkle of magic
One of the features of the application is its ability to recognize user's preferences, learn them along the way and use them when suggesting possible routes. We utilized decision support mechanisms and algorithms to facilitate these features. Application was also integrated with natural language processing engine which allows to navigate though the application using your voice.