Inspiration
We spend countless hours browsing websites for easy answers, and students are always strapped on time. Mario is here to help (starting with Drexel University) by providing crafty answers generated through speech and text analysis.
What it does
Mario is Drexel's magical assistant. He can be reached through our website online or by phone to help a Drexel student with appointment requests, room reservations, general questions, and much more!
How I built it
We used a serverless tech stack built upon ReactJS along with Google Cloud. For Mario's telephone capabilities, his speech was synthesized through Google Cloud's extensive libraries and user input was converted to text, then analyzed by Google's Dialogflow implementation.
Challenges I ran into
Training a linguistic AI isn't easy! We spent countless hours teaching Mario how to respond to certain general questions about Drexel and specific administrative requests.
Our team consists of two people: only one with programming knowledge. This was challenging at first, but through this rigorous process, he was able to pick up interest in coding and machine learning.
Accomplishments that we are proud of
Mario is a great resource for current and prospective students. We are extremely excited to show off his skills and capabilities to our peers!
What we learned
We came into this hackathon with no machine learning knowledge but was able to implement and learn on the spot. We also figured out how to implement animations in React which was a cool task!
What's next for DragonDesk
We plan to implement more functionality for Mario. Our main focus will be to train Mario to sift through information and provide more accurate and relevant answers.
It would be awesome if Drexel were to adopt Mario as a campus resource. It would be a dream come true to implement this on university campuses around the world!
We will definitely be showing him off on Drexel campus, and look forward to critical feedback from our peers!
Log in or sign up for Devpost to join the conversation.