README.md

metadata

title: AI Support Ticket
emoji: 🎫
colorFrom: blue
colorTo: purple
sdk: docker
pinned: true

AI Customer Support Ticket Resolution (OpenEnv)

An AI-powered support ticket system using AWS Bedrock for automatic ticket classification and resolution suggestions.

Built by Jainam H. Maru — Cybersecurity Engineer & AI Security Researcher

Features

• AI Classification — Automatically categorizes tickets (billing, technical, account, shipping, general)
• Priority Detection — AI assigns priority (low / medium / high / critical)
• Resolution Suggestions — AI generates resolution drafts using AWS Bedrock (Titan)
• Confidence Scoring — High-confidence tickets auto-resolve; low-confidence tickets route to agent review
• Agent Workflow — Agents can review, add notes, override AI, and close tickets
• Live Stats Dashboard — Real-time metrics on resolution rates and ticket distribution

Architecture

Customer → Submit Ticket → AI Engine (AWS Bedrock)
                               ↓
                    Classify + Generate Resolution
                               ↓
              confidence ≥ 0.8 → ai_resolved
              confidence < 0.8 → agent_review
                               ↓
                    Agent Reviews & Closes

Setup

1. Install dependencies

pip install -r requirements.txt

2. Configure environment

cp .env.example .env
# Edit .env with your AWS credentials

3. AWS Bedrock Setup

• Enable Amazon Titan Text Express model in AWS Bedrock console
• Ensure your IAM user/role has bedrock:InvokeModel permission

4. Run

python main.py

Open http://localhost:8000

API Endpoints

Method	Endpoint	Description
POST	/tickets	Submit new ticket (triggers AI analysis)
GET	/tickets	List tickets (filter by status/priority)
GET	/tickets/{id}	Get ticket details
PATCH	/tickets/{id}	Update ticket (agent actions)
DELETE	/tickets/{id}	Delete ticket
GET	/stats	Dashboard statistics

Project Structure

├── main.py          # FastAPI app & routes
├── ai_engine.py     # AWS Bedrock AI classification & resolution
├── models.py        # Pydantic data models
├── index.html       # Frontend UI
├── tickets.json     # Ticket storage
├── .env.example     # Environment config template
└── requirements.txt