Winner of the IBM "Best Use of Bluemix" award and the MLH "Hack Harassment" award!


Chris often works with international students and came up with this idea after trying to communicate with students who did not speak English well. As an international student, Jiayang shares this perspective. In real life, an interpreter would be employed to facilitate understanding. Why doesn't one exist inside of messaging apps?

What it does

TransLIT is a bot that can connect to a Slack channel. Using simple keyword commands, TransLIT sends translations of messages sent within the Slack chat, much like a real life interpreter. Through the power of the IBM Watson translation engine, conversations between people of different tounges can happen with ease.

How we built it

TransLIT is built using Bluemix, the IBM Infrastructure as a Service platform. The code -- a mix of Javascript and NodeRed -- integrates seamlessly with the IBM Watson translation service. Through a webhook, the TransLIT application is able to send and receive messages when triggered by message commands.

Challenges we ran into

We first attempted to write an application using a Javascript broker and Java backend, and were originally targeting Facebook Messenger. It turned out to be much more difficult than we first expected. After writing most of an application, we found out that Facebook had recently changed its bot API and now requires a SSL certificate, which we were not able to obtain. We then decided to switch gears and develop for Slack, which had an API which was much more reasonable to work with. We also decided to switch over to NodeRed at the same time.

Accomplishments that we're proud of

We started from scratch, and we had to restart a few times, but we ended up making a project that we find useful. We also met each other on the first day of HackFSU, and came together as a team.

What we learned

Check to see if you need an SSL cert before you build 5 Java classes.

What's next for TransLIT

We would like to build it into a full Slack app, branch out to other platforms, and store user input to make keyboard commands easier.

Share this project: