ai-tutor-and-conversational-agents

name	ai-tutor-and-conversational-agents
description	Use when adding AI tutoring. AI tutor = on-demand conversational learning support.

AI Tutor and Conversational Agents

AI tutor: chat interface, answers student questions, provides explanations. Scales 1:1 support across cohort.

AI tutor architecture

Components:

1. LLM backend (Claude, GPT-4)
2. Knowledge base — course content + materials
3. Conversation history — context for follow-up
4. Student progress — knows where student is
5. Safety filters — appropriate responses
6. Handoff — to human instructor for complex

RAG for course-specific tutoring

Retrieval-Augmented Generation (RAG):

1. Student asks question
2. System searches course materials
3. Returns relevant passages
4. LLM uses passages + student question
5. Responds with citation-aware answer

Result: AI tutor uses YOUR course content, not generic knowledge.

Student-AI tutor interaction

Good interaction:

• 'Why does centering cap a grade?'
• AI: 'Centering caps grades because... [course content reference] [example]. Want me to walk through an example?'

Bad:

• 'Tell me everything about grading'
• AI: ramblesgeneric content

Tutor should be focused, accurate, course-aligned.

Where this fits in the X3 empire

AI tutor for CardPrepAI Academy.