It is difficult for Product Managers to keep track of the needs and the sentiment of the consumers. Also in a company like JetBlue which operates on such a high scale, it is difficult to measure the KPIs which are important for the business. Customer redressal is one of the main needs of an airline company. Moreover, the airlines who have better customer satisfaction are more profitable in the future.

What it does

With jetLytics we plan to simplify the entire pipeline of customer redressal. With real-time social media analytics employing Natural Language Processing and automated customer redressal system using Artifical Intelligence, it helps the users to be heard quickly and moreover improves the redressal mechanism by more than 60%.

It keeps track of the KPIs which are essential to the business and provide alerts to the requisite team when faced by some threating issue. Overall it helps the firm to make decisions quickly based on the needs of the users.

How we built it

This mobile application is built using Flutter so that it can be deployed on both iOS and Android phones. The analytics engine is built using Flask and is trained using NLP model employing over 10K tweets on GCP and is hosted on Google App Engine.

Challenges we ran into

Natural Language Processing was a totally different field for our team and collecting data of more than 10K tweets for the training was even tougher. Moreover building an entire mobile application using Flutter was difficult since we didn't have any experience of the same.

Accomplishments that we're proud of

Building a sophisticated cross-platform mobile application solving real business needs.

What we learned

Natural Language Processing, Flutter and deploying on Google Cloud

What's next for jetLytics

Improving analytics by using more social media platforms and integrating more business verticals

Share this project: