Traditionally, customer feedback is static and only comes into play when there are complaints. What about retrieving intuitive and authentic passenger opinions and feelings instead?

What it does

With happytrains, BVG passengers are motivated to engage in a feedback process by earning points for high quality answers. The feedback is sent while the customer is still in the vehicle -- this can be achieved by sending them a push notification in the middle of their ride, for example depending on the last connection they searched for or by fetching their location.

Due to the gamification mechanism, users give high quality feedback that is then analysed to extract the feelings that are conveyed. happytrains then calculates the average feelings customers have when they ride a train and provides a dashboard to BVG.

For now, this dashboard provides the feelings for each trainle and shows the passengers' reasons for feeling the way they do -- but it is easy extensible with more vehicles and deeper analysis.

How we built it

The passenger interface is built with a mobile angular.js application that provides the behaviour that might also be implemented by the BVG app itself. happytrain's backend is written in Python and IBM Watson Tone Analyser is used for sentiment analysis of customer feedback. All data is managed with GraphQL. The BVG dashboard is a dynamic web app that works with jQuery.

Challenges we ran into

A major hurdle we encountered early on was finding a good idea, i.e. one that we all found exciting, useful and feasible. Especially in later stages refocussing on essential task proved to be difficult.

Accomplishments that we're proud of

From the beginning our goal was not only realising our idea but simultaneously getting used to new technologies that we had not dealt with before. By essentially integrating both IBM Watson and GraphQL we challenged ourselves with two such new technologies that we mastered enough to make them part of our project. Also: We made it work!

What we learned

Many ideas might seem great at first, but soon reveal deficiencies when look at more closely. This insight, while frustrating at first, is one of our major take-aways. More importantly, however, not giving up when a good idea seemed impossible to find, was an invaluable lesson to learn.

What's next for happytrains

We'd be happy to be able to show what customer feedback can look like when done right!

Share this project: