Inspiration
While computers are a very common household item, most people don't understand computers and the parts that make them up. For beginners looking to upgrade a computer or build a new one, it can be an extremely daunting challenge. Building a computer for the first time can require hours of research just to understand what the computer requires to run. We wanted to streamline this process and make it easier for anyone to learn about computers just by asking questions.
What it does
We created a talking AI that can communicate with the user in a conversational manner, explaining how certain parts work and why they are important. It can define certain parts, compare different manufacturing companies, and even advise the user on safety practices.
How we built it
We created the product using AIML, an XML derivative that stands for Artificial Intelligence Markup Language. We had to define hundreds of categories, sentence templates, and language maps in order to ensure we had a wide range of sentences the AI could understand. We then used Node.js to compile and create a project using Heroku app.
Challenges we ran into
Our biggest challenge was finding a way to compile our AIML database into a working application. We initially tried to incorporate the API directly into an HTML webpage, but thanks to the help of Steve we were able to use Node.js to simplify the process and get the website up and running.
Accomplishments that we're proud of
Versatility of our program - given the time we had, we were able to make a system that accepts a wide range of input. Teamwork - we split the workload evenly and were each assigned a task so that each person had an opportunity to contribute. Organization & Planning - Before beginning the project we planned out the majority of the project on our various whiteboards and discussed our strategy of how to code past various input challenges.
What we learned
The entire team learned how to use AIML well enough to write a similar program in the future. We learned how to use Node.js and Heroku app in conjunction
What's next for Sleeper-CellHelper
Incorporate machine learning - as we created Sleeper-CellHelper we learned about the learnf function, which will allow us to teach our AI by talking to it during the beta phase. This would greatly streamline the development process and allow us to expand our team even to those who aren't interested in coding. Better website and mobile app - our current page is only a placeholder as we plan to add a better UI and create a companion android app with the same function as the website.
Note: Before testing our product, please bear in mind it is a work in progress and is still somewhat limited in scope. Try sentences like "What is a solid state drive," "Compare Intel and AMD processors," or "How can I build a computer?" Feel free to use alternative sentence structures.
Log in or sign up for Devpost to join the conversation.