Peppa-Bot

Challenge: The Career Center has adopted #CarlsCreate as its official slogan, and we want our website to reflect this vision! What resources should be available for Carls to integrate and apply their liberal arts education toward succeeding in their career paths and live a meaningful life after Carleton? How should the site be organized for students to find that information best? How can new technologies be implemented to increase engagement and improve user experience on the Career Center website?

Inspiration

Chatbots aren't a new concept and are pretty well integrated into most small businesses strategies. They're so widespread that:

  • By 2020, 80% of businesses will have some sort of chatbot automation in place (Business Insider, 2016).
  • 21% of consumers see chatbots as the easiest way to contact a business (ubisend, 2017).

That's the scene with small business, but what about College Campuses? Here too we're seeing an emergence of chatbots available on websites to assist students in their day to day functions and uncover resources that students find difficult to find! Universities all around the world are responding to increasing demands from their student body by automating the "resource discovery" process. Being able to answer student questions 24x7 without driving staff into the ground is an invaluable resource, and that's where Peppa-Bot comes in!

Instead of spending our energies in answering questions that have answers available on our site, we can focus on the counseling aspects of being at the Career Center. Additionally, we're making the information we do have to provide more accessible to students that can't make it to the office.

What It Does

Peppa Bot gets started with a simple "Hey!". Peppa can help with Application Materials, Majors and Careers, Internships, Life after Carleton, and Networking. Each of those categories has its level of detail, but the overall idea is that it'll help you figure out what resource to use. Currently, it doesn't do a great job of supporting text inputted by the user, but the suggestion text works through and through.

How I Built It

Chatfuel powers this chatbot!

Challenges I Ran Into

I started this project of building and hosting my custom webhook, working in combination with Dialogflow. I soon realized two things: it's incredibly complicated to set the whole pipeline up and would be hard for people with no CS background to maintain. So I turned to logic-based chatbot platforms that do all the backend and maintenance for me and require me to focus on what the available commands need to be.

Figuring out what parts of the site needed to be accommodated was the next step.

What's next

I'm working on incorporating more NLP portions into this project. The next step is to set up more phrases that Peppa-bot will understand and eventually move it back to Dialogeflow and build a fully custom backend for the project.

Team

Ritvik Kar  ·  Senior, Computer Science and Cinema & Media Studies Major

Built With

  • chatbot
  • chatfuel
Share this project:

Updates