Inspiration A lot of companies want to determine if their employees are happy at work, but do not know how to calculate this happiness. Some tools offer solutions, but require a user interaction. Some people might not give their actual thoughts when feeling they are forced to give a feedback.

What it does Monitors channels that it was added to by analyzing the messages that are sent in that channel. It determines whether the message was positive or negative and extracts the topic of that message.

How I built it We built a bot that analyzes the messages from channels it is subscribed to and stores the result in a cloud database. A dashboard allows to determine the feelings on every channel for a certain period of time.

Challenges I ran into Finding a proper sentiment analysis engine with support for tags in messages.

Accomplishments that I'm proud of A pretty unique and creative solution to a challenging issue. Eve can interact with the users to obtain information while providing an easy to understand overview of the data.

What I learned Machine learning is not always precise. The Slack API is pretty easy to use and understand.

What's next for Eve Configurable triggers when the happiness of a channel reaches a certain level. Possibly, the switch to a more precise machine learning API. Be production ready.

Built With

Share this project:

Updates