The need for a faster and more comfortable way to obtain information about student loans. We wanted to make it as easy to access and use as possible. That's why we chose to have our chat bot on Facebook Messenger, eliminating the need to install a separate app. It also works on Skype, just to be sure it is as easy to reach as possible! We also started working on a website to handle queries using voice recognition. Right now it recognizes voice and outputs it to text. This can now be passed to our bot to handle the rest.

What it does

We have trained the bot to answer a number of questions about student loans. Application process, tuition fees, how you will repay the loan etc. Apart from that, If you still have questions, after asking you where you're from, it can put you in touch with a student from your country who has experience with student loans so he can give you some tips.

How we built it

We used node.js for our scripts. We created a bot using the Microsoft bot framework. Then, using the messenger platform for developers, we connected it to our Facebook page. Now the input from the user would be passed by the bot framework to our server, hosted using ngrok. The input was passed to our RECAST.AI bot, which we trained to recognize multiple intents. We made it recognize the key words, as well as more natural language (e.g "uni" or "postgrad"). This would provide another layer of comfort for the user. After the input is interpreted, the appropriate answer is sent to the user.

Challenges we ran into

None of us had done anything like this before. We had no idea how any of the services we used worked, or even what the syntax of node.js was.

Accomplishments that we're proud of

We made a bot! It talked back, it answered our questions! We managed to understand new technology quickly, and most importantly understand what else we need to learn in order to make better bots.

What we learned

Apart from the languages and individual services we used for the first time, we managed to learn how to use them all at once. We learned that connecting services to a script and one another is not as tough as it looks.

What's next for SLC: The Bot.

The bot's intents are implemented in separate files. This means new things can be added quickly, and parts can be updated. In the future, we could make the bot understand language even better, add new things it can answer or connect it to a database of information regarding your personal student loan. The data provided by also has many other components, like sentiment analysis, which could be used.

Built With

Share this project: