NMKR Support AI Agent

An intelligent AI-powered support system for NMKR, designed to provide automated, accurate responses to user queries about NMKR's products and services. This system uses CrewAI to orchestrate multiple AI agents that work together to provide comprehensive support responses.

Overview

This project implements an AI support system that:

• Processes support requests through REST API and webhooks
• Automatically crawls and analyzes NMKR documentation
• Provides detailed, context-aware responses about NMKR services
• Handles asynchronous processing with Redis queue
• Integrates with Plain for customer support workflows

Prerequisites

• Docker and Docker Compose
• Python 3.10 or higher
• OpenAI API key
• Plain webhook secret (for webhook integration)

Quick Start

1. Clone the repository:

git clone <repository-url>
cd nmkr_support_v4

2. Create a .env file with your credentials:

MODEL=gpt-4o
OPENAI_API_KEY=your_openai_api_key_here
WEBHOOK_SECRET=your_webhook_secret_here
ANTHROPIC_API_KEY=your_anthropic_key_here  # Optional
SPIDER_API_KEY=your_spider_key_here        # Optional
FIRECRAWL_API_KEY=your_firecrawl_key_here # Optional

3. Start using the convenience script:

chmod +x start.sh
./start.sh

Or manually with Docker Compose:

# Build and start services
docker-compose up --build -d

# View logs
docker-compose logs -f

# Stop services
docker-compose down

API Endpoints

Submit Support Request

curl -X POST "http://localhost:8000/api/support" \
     -H "Content-Type: application/json" \
     -d '{"query": "How much does it cost to do an Airdrop with NMKR?"}'

Response:

{
    "job_id": "123-456-789",
    "status": "queued"
}

Check Request Status

curl "http://localhost:8000/api/support/status/123-456-789"

Plain Webhook Integration

curl -X POST "http://localhost:8000/api/webhook" \
     -H "Content-Type: application/json" \
     -H "Plain-Workspace-Id: ws_123" \
     -H "Plain-Event-Type: thread.created" \
     -H "Plain-Event-Id: evt_123" \
     -H "Plain-Signature: your-webhook-signature" \
     -d '{
       "id": "evt_123",
       "type": "thread.created",
       "payload": {
         "message": {
           "content": "How much does it cost to do an Airdrop with NMKR?"
         }
       }
     }'

Project Structure

nmkr_support_v4/
├── src/
│   └── nmkr_support_v4/
│       ├── api.py                         # FastAPI application
│       ├── crew.py                        # CrewAI configuration
│       ├── queue_manager.py               # Redis queue management
│       ├── tools/
│       │   └── custom_tool.py            # Web crawling tools
│       ├── links_with_descriptions.json   # NMKR links data
│       └── docs_links_with_descriptions.json
├── docker-compose.yml                     # Docker services configuration
├── Dockerfile                            # Container build instructions
├── requirements.txt                      # Python dependencies
├── setup.py                             # Package configuration
└── start.sh                             # Convenience startup script

Development

1. Install in development mode:

pip install -e .

2. Run locally without Docker:

# Start Redis
redis-server

# Start RQ worker
rq worker nmkr_support

# Start API
uvicorn nmkr_support_v4.api:app --reload

Configuration

Environment Variables

• MODEL: OpenAI model to use (default: gpt-4o)
• OPENAI_API_KEY: Your OpenAI API key
• WEBHOOK_SECRET: Secret for Plain webhook verification
• REDIS_URL: Redis connection URL (default: redis://localhost:6379)

Docker Services

• API: FastAPI application serving endpoints
• Worker: RQ worker processing support requests
• Redis: Queue and cache management

Useful Links

• NMKR Website
• NMKR Documentation
• NMKR Developer Portal
• CrewAI Documentation

Contributing

1. Fork the repository
2. Create a feature branch
3. Commit your changes
4. Push to the branch
5. Create a Pull Request

License

[Your chosen license]