Tutor-multi-ai-agent

# Tutor Agent: Multi-Agent AI Tutoring System

## Table of Contents

1. [Project Overview](#project-overview)
2. [Live Demo](#live-demo)
3. [Screenshot](#screenshot)
4. [Key Features](#key-features)
5. [Architecture](#architecture)
6. [Directory Structure & Index](#directory-structure--index)
7. [Setup & Installation](#setup--installation)
8. [Configuration & Environment](#configuration--environment)
9. [Usage Guide](#usage-guide)
10. [Development Playbook](#development-playbook)
11. [Deployment](#deployment)
12. [Best Practices](#best-practices)
13. [Contributing](#contributing)
14. [FAQ](#faq)
15. [Challenges & Solutions](#challenges--solutions)
16. [References & Further Reading](#references--further-reading)

---

## Project Overview

**Tutor Agent** is an AI-powered educational assistant that leverages a multi-agent along with multi turn architecture to provide specialized tutoring in mathematics and physics (with extensibility for more domains). It uses Google's Gemini API for LLM-powered reasoning, and routes student queries to the most appropriate specialist agent, which can use tools like calculators and knowledge bases to generate comprehensive, step-by-step answers.

---

## Key Features


### Smart Routing

- Automatic subject detection
- Context-aware agent selection
- Seamless agent switching when needed
- Multi-agent collaboration for complex queries

### Specialist Agents

- Math Agent for mathematical problems
- Physics Agent for physics concepts
- Extensible architecture for new domains
- Tool-equipped agents for specialized tasks

### Multi-Turn Conversations

- Maintains context across multiple interactions
- Builds on previous questions and answers
- Allows follow-up questions and clarifications
- Persistent conversation history with MongoDB

### Rate Limiting

- Protects API endpoints from abuse
- Configurable limits per user/IP
- Redis-based rate limiting implementation
- Graceful handling of limit exceeded cases

### Caching with Redis

- Fast