StudyPeerAI

StudyPeer AI matches students by strengths, schedules group study, and provides an AI tutor for explanations, quizzes, and summaries — helping you study smarter together.

Comment

Inspiration

Students often study alone or in poorly matched groups where strengths and weaknesses don’t complement each other. This creates inefficient learning and low motivation. We wanted to build a platform that makes collaboration easy, productive, and smart. The idea behind StudyPeer AI came from wanting students to learn together with structured support and AI guidance — so no one feels left behind.

How we built it

We built StudyPeerAI using React for the UI and Firebase for real-time data, authentication, and group scheduling. The AI layer (explanations, quizzes, summaries) is powered by the OpenAI API.

For group formation, we implemented a lightweight mathematical matching system. Each candidate group receives a Final Match Score computed using a weighted multi-criteria model:

$$ \text{FinalScore} = \sum_{i=1}^{n} (C_i \times W_i), \quad \sum W_i = 1 $$

Expanded:

$$ \text{FinalScore} = 0.30,S_{\text{subject}} + 0.25,S_{\text{skills}} + 0.20,S_{\text{academic}} + 0.10,S_{\text{year}} + 0.10,S_{\text{location}} + 0.05,S_{\text{institution}} $$

Each sub-score is normalized: ( C_i \in [0, 1] )

Example sub-scores

Academic level: $$ S_{\text{academic}} = \max!\left(0,;1 - \frac{|L_{user} - L_{avg}|}{4}\right) $$

Skill complementarity: $$ S_{\text{skills}} = \min!\left(1,\frac{M_{help} + 0.8,M_{teach}}{T_{possible}}\right) $$

Subject match: $$ S_{\text{subject}} = \min!\left(1,,0.8D + 0.4C + 0.6M + 0.1I\right) $$

Similarity measures

Jaccard similarity: $$ J(A,B) = \frac{|A \cap B|}{|A \cup B|} $$

Cosine similarity: $$ \cos\theta = \frac{A \cdot B}{|A| |B|} $$

Skill proficiency

$$ P = \min(100,;10 + 15K) $$

Score normalization

$$ S' = \frac{S - S_{min}}{S_{max} - S_{min}} $$

This combined mathematical system ensures fast, fair, and scalable group recommendations.

Challenges we ran into

  • Designing accurate group-matching from limited user input
  • Ensuring AI explanations were clear and contextually relevant
  • Managing real-time collaboration and scheduling
  • Maintaining clean UI under time constraints

Accomplishments that we're proud of

  • Built a functional peer-matching + AI study assistant system
  • Successfully integrated AI to generate helpful quizzes and summaries
  • Created an intuitive interface that makes group learning seamless
  • Designed a workflow that motivates collaboration rather than solo learning

What we learned

  • Working with real-time databases (Firebase)
  • Building matching algorithms with dynamic inputs
  • Prompt engineering for educational use-cases
  • Designing simple systems that encourage collaboration
  • How AI can amplify human learning rather than replace it

What's next for StudyPeerAI

  • Add video + voice study rooms
  • Introduce topic-level analytics and personalized learning paths
  • Expand to cross-college matching
  • AI-generated study plans based on goals and deadlines
  • Mobile app for easier access
  • Add gamification: streaks, XP, badges
Share this project:

Updates