The proposed solution is a structured AI-Sandbox & Analytics Platform integrated with 
AI-literacy courses and project-based assignments. It records how students interact with 
AI tools, prompts reflection, and provides analytics to guide both learners and 
instructors. The platform includes a sandbox workspace for safe experimentation, a 
scaffolding layer offering guidance on prompting, source evaluation, and accuracy 
checking, and micro-modules addressing bias, hallucinations, and responsible AI use. 
Instructors access an analytics dashboard that highlights usage patterns, learning gaps, 
over-reliance on AI, and suggests targeted interventions. Key technologies include an 
LLM-based agentic workspace, logging layer, analytics dashboard, scaffolding engine, 
safe data storage, machine-learning patterns, and versioning tools.
The platform fosters active project learning by allowing students to experiment with 
prompts, receive feedback, and develop AI literacy gradually, positioning AI as a learning 
scaffold rather than a shortcut. Success can be measured through improvements in
prompting, verification, and reflection skills, reduced academic misconduct, stronger 
project outcomes, and instructor confidence. At the institutional level, aggregated 
analytics can track year-over-year improvements in AI literacy and responsible AI use
among students.