Open Ticket AI

Quick Start Guide ​

Get started with Open Ticket AI in 5 minutes.

Prerequisites ​

  • Python 3.13+
  • Access to an OTOBO, Znuny, or OTRS instance
  • API token or credentials for your ticket system

Installation ​

Install Core Package ​

bash
# Using uv (recommended)
uv pip install open-ticket-ai

# Or using pip
pip install open-ticket-ai

Install Plugins ​

bash
# Install OTOBO/Znuny plugin
uv pip install otai-otobo-znuny

# Install HuggingFace plugin (for ML)
uv pip install otai-hf-local

# Or install the complete bundle
uv pip install open-ticket-ai[all]

First Configuration ​

1. Set Environment Variables ​

bash
export OTOBO_BASE_URL="https://your-ticket-system.com"
export OTOBO_API_TOKEN="your-api-token"

2. Create Configuration File ​

Create config.yml:

yaml
# Load plugins
plugins:
  - name: otobo_znuny
    config:
      base_url: "${OTOBO_BASE_URL}"
      api_token: "${OTOBO_API_TOKEN}"

# Configure pipeline
orchestrator:
  pipelines:
    - name: classify_tickets
      run_every_milli_seconds: 60000  # Run every 60 seconds
      pipes:
        # Fetch open tickets
        - pipe_name: fetch_tickets
          search:
            StateType: "Open"
            limit: 10

        # Log tickets (for testing)
        - pipe_name: log_tickets

Running Your First Pipeline ​

bash
# Run the pipeline
open-ticket-ai run --config config.yml

# Or with verbose logging
open-ticket-ai run --config config.yml --log-level DEBUG

You should see output like:

[INFO] Loading configuration from config.yml
[INFO] Initializing plugins...
[INFO] Starting orchestrator...
[INFO] Running pipeline: classify_tickets
[INFO] Fetched 10 tickets
[INFO] Pipeline completed successfully

Next Steps ​

Add Classification ​

Update your config to classify tickets:

yaml
orchestrator:
  pipelines:
    - name: classify_tickets
      run_every_milli_seconds: 60000
      pipes:
        - pipe_name: fetch_tickets
          search:
            StateType: "Open"
            limit: 10

        # Add ML classification
        - pipe_name: classify_queue
          model_name: "bert-base-uncased"

        # Update tickets
        - pipe_name: update_ticket
          fields:
            QueueID: "{{ context.predicted_queue_id }}"

Explore Examples ​

Check out complete examples:

bash
# List available configExamples
ls docs/raw_en_docs/config_examples/

# Try the queue classification example
cp docs/raw_en_docs/config_examples/queue_classification.yml config.yml
open-ticket-ai run --config config.yml

Learn More ​

Common Issues ​

Connection Error ​

Error: Failed to connect to ticket system

Solution: Verify OTOBO_BASE_URL is correct and accessible.

Authentication Error ​

Error: 401 Unauthorized

Solution: Check that OTOBO_API_TOKEN is valid and has required permissions.

Plugin Not Found ​

Error: Plugin 'otobo_znuny' not found

Solution: Install the plugin:

bash
uv pip install otai-otobo-znuny

Getting Help ​

What's Next? ​

Now that you have Open Ticket AI running:

  1. Customize Configuration: Adapt to your workflow
  2. Add More Pipes: Enhance functionality
  3. Monitor Performance: Track classification accuracy
  4. Scale Up: Process more tickets
  5. Contribute: Share your experience and improvements