Customer Support AI Agent

Overview

A sophisticated AI-powered customer support system built using the PHIDATA Agent Framework. The system uses historical support ticket data and documentation to provide accurate and contextual responses to customer inquiries.

Features

• Multiple specialized AI agents for different support scenarios
• Training based on real support ticket data history
• Documentation-aware responses
• Code-related query support
• Interactive playground environment

AI Agents

1. Documentation Agent

• Specializes in Materio VueJS admin template documentation
• Provides detailed responses for documentation-related queries
• Trained on comprehensive documentation dataset

2. Support Ticket Agent

• Handles queries based on historical support tickets
• Trained on resolved support ticket database
• Leverages past solutions for current problems

3. Customer Support Agent

• A composite agent combining Documentation and Support Ticket agents
• Provides comprehensive support coverage
• Intelligent routing between specialized agents

4. Code Scraper Agent

• Specialized in code-related queries
• Trained on Materio VueJS admin template codebase
• Assists with technical implementation questions

Data Sources

Support Ticket Data

• Location: support_ticket_data/ directory
• Content: Historical support tickets with questions and answers
• Collection Method: Automated via scripts/github_issue_scraper.py

Documentation Data

• Comprehensive Materio VueJS admin template documentation
• Structured for AI training and reference
• We have provided a links of the documentation in the codebase for the documentation

Code Data

• Location: data/ directory
• Source: GitIngest collection
• Scope: Materio VueJS admin template codebase

Getting Started

Basic Usage

Run a specific agent:

python [agent_name].py

Interactive Playground Environment

Launch the playground environment:

python playground.py