Smart Solution

Inspiration

Just to get a better view of the current situation, we called up the customer service and we disconnected after 7 min of wait time.

What it does

How I built it

Based on the user actions in the existing app and with the power of machine learning, we built a prediction model for what issue the user might have contacted for and also encourage the user to use the chat service as customer representatives can handle multiple chats at a time but not calls.

Challenges I ran into

Analysing the user actions and training the model created by us was a challenge as we did not have large data set to begin with. But our solution will give a data set over time and with customer service team and our report model we can train it attain high accuracy.

Accomplishments that I'm proud of

We were able to analyse the user actions and we have a pretty good model with acceptable accuracy with small dataset that we could generate in time.

What I learned

Team work, how to handle pressure and basic concepts of machine learning.

What's next for Smart Solution

It's integration friendly and we look to increase the accuracy of the prediction and also the metrics data could also benefit T mobile from marketing and promotion point of view.

Built With

• angular.js