Trava - A trip-planning app with a social network feature

The ClientTrava is a startup company in the United States that provides a trip-planning social app.

Goal

Trip planning might be stressful and confusing as the trip schedule might only meet some team members' expectations. Trava wanted to build an app that made trip planning more accessible by suggesting what to see and what activities to take based on user's preferences.

Need

The client needed a mobile app that allows users to invite friends and family, create a trip and vote for the activities and places they would like to visit. Based on that, the Trava app was to provide a logistically optimised itinerary selecting a specified number of attractions. To do that, the client needed a Machine Learning algorithm so that users could vote for a few items, but the app should create an itinerary based on all available options.

Trava also needed an admin panel to manage users, directions and attractions. The system should gather information from globetrotters and travellers, but users should be able to add their items too.

Challenges

To the challenges, we would include a front-end based on custom designs and sophisticated data management. As destinations and attractions can be added to the calendar or favourites or be voted on, managing those in case of changing or deleting was needed. We had to make sure that changes won't affect items observed by users.

To provide the itinerary functionality, we had to build a Machine Learning algorithm that works well on small data sets. Users' opinions expressed in voting are just a sample for the algorithm to establish their preferences and build the itinerary based on them. We also had to build a social network functionality, which can be considered a challenge. Users can add stories from their trips that contain images and videos, and those can be observed by users, commented and liked. Those posts should also be recommended to users interested in such content. Our solutionWe've built a mobile and web app from scratch. We dynamically conceptualised many aspects of the project hand in hand with the client. We used the TeaRex.AI recommendation engine to offer relevant posts to users.

Process and Cooperation

Although the client provided designs, we started the project by reviewing them and defining the final flow and look of the app.

Having those, our development team received precise requirements and a few months after, the first version of the app was ready to be reviewed by the client. Since then, we have worked in two-week sprints, defining the current scope in cooperation with the client.

We knew the primary goal, but the client could flexibly modify the scope in accordance with the needs.

We've had a significant impact on the project as we conceptualised the project together and advised our client on how to develop the app to receive the ultimate value for the same price.

Results

In a year, we've built an MVP app that provides many features but still needs optimisation.