Timothy Birmingham

CARL: Classroom Assistant using Real-time LLMs

5th Place Winner, 2024 Nittany AI Challenge ($2,000)

The Problem

K-12 students exhibit 55% higher engagement levels in group activities compared to traditional lectures. However, teachers are compelled to allocate 54% of their time towards whole-class lectures due to time and resource constraints when addressing smaller groups.

  • Limited instructional support in group learning
  • "One Size Fits All" approach for different learning styles
  • Stale classroom discussions

Our Solution

CARL is an AI teaching assistant designed to foster dynamic interactions between students and teachers, offering tailored support for active learning strategies, such as group activities. In turn, CARL enhances classroom engagement, provides personalized learning, and promotes critical thinking.

How CARL Works

  • Guides group activities based on custom instructions
  • Stores classroom discussion and whiteboard state
  • Responds to student's text and visual questions instantly
  • Generates warm-up and exit questions for recall

Team

Jared Daniel, Andrew Klawa, Neha Pandit, Timothy Birmingham, Sunghee Choi

Faculty Advisor: Robert Avanzato

Mentor: Leslie Wubbel

Coach: Joel Seidel