ThothBot

Inspiration

It all started one innocent weekday afternoon (well, all weekday afternoons), when I found myself woefully unprepared for tests and quizzes that would inevitably plague me the following period. Of course, by the time these sinister exams had entered by shamefully inadequate attention span, it was often far too late for me: I often had less than thirty minutes to study, and, like the famed successor of house Snow, I knew nothing. Perhaps I could pull out some cheat sheet or textbook and scrape together enough knowledge to fend off the dreaded D or F, but often my time was up barely after I opened my book. Finally catching wise of my self-destructive habits, I made a strong proactive effort towards positive change.

No, just kidding. I made a chatbot that let me study faster.

What it does

ThothBot takes advantage of the Quizlet API for a studying experience with the effectiveness of Quizlet yet the speed of messaging, allowing users to study effectively and quickly while on the go.

How we built it

Using a personal server, we accessed the Quizlet API as well as our Gupshup chatbot to provide users access to their personal Quizlet sets. We coded the chatbot to test the users on their sets in a lightweight and portable way on Facebook Messenger.

Challenges we ran into

Authentication was a huge challenge for us as we struggled to allow our chatbot to send the proper authentication tokens to the server and to Quizlet, which then had to send back tokens that would differentiate between users. This was compounded by Gupshup's rather limited database commands.

Accomplishments that we're proud of

We're extremely proud of how far we've come integrating the different parts of this chatbot considering we had very little experience in almost all of these parts (mentors were a great resource, of course). Also, it works!

What we learned

As our first exposure to Gupshup and chatbots in general, ThothBot helped us to understand how chatbots function and, of course, how to construct one. In addition, we learned a huge amount about javascript, GET and POST commands, and the JSON, all of which we had limited experiences with in the past.

What's next for ThothBot

Machine learning and natural language processing could advance our chatbot significantly, allowing it to learn and adapt to different users. We could extend integration into Skype and SMS, taking advantage of the uniquesnesses of those features along the way. Finally, for much greater studying flexibility, we could integrate with other highly regarded studying services such as GoConqr, Evernote, and StudyBlue.

Built With

• chatbot
• gupshup
• javascript
• json
• messenger
• php