Challenge

Obtaining sufficient on-the-job experience for teaching professionals is a slow process. According to estimates, 5000 hours of teaching practice equate to 5-6 years of teaching experience However, these experiences are often unintentional, lacking feedback from qualified mentors and repetitive practice in teaching specific material to students with varying backgrounds or learning difficulties. Unlike many other professions, gaining experience through trial and error can come at a significant cost for students in classes taught by less experienced teachers.

RnD Objectives

We propose creating an open teaching simulation platform that leverages recent advancements in generative AI and agentic workflows. This platform simulates group dynamics and provides expert feedback. Within this virtual classroom, teachers can practice with students who are simulated by a collection of Large Language Model (LLM) agents. These agents are primed with specific behavioral traits, learning patterns, and prior knowledge. Additionally, the platform includes “mentor LLM agents” that monitor class interactions and offer personalized feedback and recommendations based on the analysis of teacher performance.

People