Inspiration

Our team's inspiration for OLGA (Optimized Learning and Guidance Assistant) came from our own experiences as students at CMU-Africa. We vividly recall the stress and confusion of course selection periods - messaging second years for advice, waiting for available slots on the academic advisor's calendar, and feeling overwhelmed by the pressure to choose courses before the registration deadlines. We realized that this challenge wasn't unique to us; it was a systemic issue affecting the entire student body. The vision of a tool that could provide personalized, round-the-clock guidance to every student, regardless of their background or career aspirations, drove us to create OLGA.

What it does

OLGA is an AI-powered academic planning assistant designed specifically for CMU-Africa students. It offers:

  • Personalized course recommendations based on student profiles, interests, and career goals
  • Real-time tracking of degree requirements and progress
  • Interactive degree planning with an intuitive, card-based interface
  • Integration of career services insights into academic planning
  • 24/7 availability for academic guidance and support

By combining these features, OLGA empowers students to make informed decisions about their academic journey, aligning their course selections with both degree requirements and long-term career aspirations.

How we built it

We haven’t fully built OLGA yet, but we developed a high-fidelity prototype using Figma. This prototype visually represents the user journey and the system's core functionalities. Our design was shaped by interviews with students and advisors, as well as insights from Georgia State University's advising model.

Although OLGA is currently in the design phase, we have a detailed vision for its development. Here's how OLGA will work:

Data Collection & Updates

OLGA will continuously scrape course information (descriptions, instructors, schedules, feedback) from CMU-Africa’s website using technologies like BeautifulSoup. This ensures students always have the most current and relevant data for academic planning.

Student Profile Creation

Students will input their academic history, interests, and career goals into the platform. OLGA will use this information to build a personalized profile for each student, enabling course recommendations tailored to their unique needs and preferences.

Tailored Course Recommendations

OLGA employs the BM-25 algorithm and an ensemble of retrieval methods to precisely match courses with students’ profiles. This approach ensures recommendations are highly relevant, aligned with both academic requirements and career aspirations.

Personalized Response Generation

Using the best-performing Large Language Model (LLM), such as ChatGPT or LLaMA, OLGA will generate personalized responses based on the student’s profile and course data. It will provide clear academic advice, aligning course selections with career goals and optimizing the learning experience.

24/7 Academic Support

OLGA will offer continuous, round-the-clock academic support. Students can access real-time recommendations and make strategic decisions about their academic path at any time, improving overall satisfaction and success.

Challenges we ran into

  • Recruiting students for the research interviews.
  • Choosing the area of solution to focus on

Accomplishments

  • Successfully creating a high-fidelity prototype that visualizes the solution.
  • Generating strong interest from students who validated the need for such a tool.

What we learned

This project reinforced the importance of user-centered design and the complexities of integrating AI into educational systems. We learned that technology should complement, not replace, human expertise, and the need for adaptability in course planning to align with students’ career aspirations.

What's next for OLGA

  • User Testing: Conducting further testing to validate the platform’s accuracy and ease of use.
  • Developing the software: Moving from prototype to full development.

Built With

  • figma
+ 46 more
Share this project:

Updates