Inspiration
Knowing how stressful high school can be, we want to provide an outlet for students who may not be able to get immediate help.
What it does
Allows user to input a statement with school-related subjects like stress, time management, etc. and then responds with a NLP statement that answers their words.
How we built it
We used Cohere's toolkit for Node.js in order to train our model and seamlessly integrate it into a web app. The model was trained on 92 sample prompts and responses we created, including information on a variety of high school courses at all levels: academic, honors, AP, and IB.
Challenges we ran into
We had initially planned on using the Python toolkit and Django in order to create the web app however we ran into problems with the way the toolkit was storing data, so we had to switch to Node.js. Due to limitations in Express.js, it was a challenge to output our chatbot's response to the website, so we had to create a workaround creating a separate EJS file to store the output.
Accomplishments that we're proud of
Solving the challenges we had while producing a clean, simple website anyone could use. Prior to this project, we didn't have much experience with Node.js, so it was certainly nice to learn it and use it in our project
What we learned
We learned to train natural language processing modules using Cohere as well as front-end and back-end development in Node.js using various libraries.
What's next for Henri Helps: Motivational Chat Bot for Students
Hopefully expand on Henri's knowledge by training on more data, providing fuller answers and aiding students more and more in the future.
Built With
- cohere
- css
- ejs
- express.js
- html
- javascript
- node.js
- render
Log in or sign up for Devpost to join the conversation.