Case 2: Smart Course Selector AI Agent - PH, JH, BH

Home Page
Student Login
Student Signup
Admin Login
Admin Signup
Admin Portal
Chatbot Page

To run a development server: In the bash, create a terminal and state 'ollama serve'. Open a second terminal and state 'python manage.py runserver'. This will run ollama so the chatbot page works when running the dev server.

Inspiration

We were inspired by the stress students face each semester while picking classes. With so many variables — prerequisites, time conflicts, credit limits, and career goals — it’s hard to know what to choose. We wanted to create something that makes this easier and more personalized.

What it does

The Smart Course Selector AI Agent helps students choose courses based on their subject preferences. Students interact with a chatbot that recommends classes, avoids conflicts, checks prerequisites, and tracks degree progress — all in one place.

How we built it

We used Django for the web framework and connected it to Ollama 3.2, a local AI model that interprets user input. Key components include:

Chatbot interface; Ollama3.2-powered scheduling assistant
Responsive front-end
Persistent chat history
Clear chat function
Logout button
Account storage; both student and admin
Different functionality depending on User type; admin can view/drop classes

Challenges we ran into

Setting up AI chatbot communication using a local model and preferences
Storing account/user data
Ollama3.2 integration
Ollama3.2 running very slow due to it being local; varies by device

Accomplishments that we're proud of

An interactive chatbot interface for a conversational experience
Clear and Logout functions on the Chatbot page
Clean UI with login/signup for both students and admins
Time zone timestamps
Clear chat box option
Admin portal to add/drop classes
Integration of historical data (a data set from archived old IU class schedule)
Interests page for students to choose academic interests before talking to the chatbot