Open Ticket AI

Open Ticket AI 1.4: The First Major Production Release ​

Open Ticket AI 1.4 is here, marking the first major production-ready release! This version brings enterprise-grade features, a mature plugin ecosystem, and the flexibility to automate your ticket system workflows like never before. Get the full release on GitHub.

https://softoft.sirv.com/open-ticket-ai/Open-Ticket-AI-Release-Version-1.png

Checkout the Demo! ​

This Demo shows the OTOBO Ticketsystem with OTAI setup to classify queue and priority based on the ticket content. You can login with the following link!

German Models!

The Queue and Priority model only work for German tickets as it was trained with German data.

OTOBO Queue Priority Demo

Example!

Subject: "DRINGEND! Wohnung in Mainzer Straße 8 Heizung kaputt;" Text: "Hallo, meine Heizung in der Wohnung in der Mainzer Straße 8 funktioniert nicht. Bitte um schnelle Hilfe! dringend! Die Heizungsrohre sind kalt und es ist sehr kalt in der Wohnung. Vielen Dank! Mit freundlichen Grüßen, Max Mustermann "

Queue Priority - Test Ticket

These are just test models. You can use any model you want with Open Ticket AI! It works with Huggingface Models

interesting models: OpenAlex/bert-base-multilingual-cased-finetuned-openalex-topic-classification-title-abstract oliverguhr/german-sentiment-bert siebert/sentiment-roberta-large-english distilbert/distilbert-base-uncased-finetuned-sst-2-english

Often it is better to train your own models with your own data. Then you need to publish this to huggingface_hub change the model and HF_TOKEN in the config.yml and restart OTAI.

What Open Ticket AI 1.4 Offers ​

Powerful Plugin Architecture ​

Install only the capabilities you need through a modular plugin system. Plugins extend Open Ticket AI with custom ticket system integrations, ML models, and processing logic—all without touching core code.

  • OTOBO/Znuny Plugin (otai-otobo-znuny): Connect to OTOBO, Znuny, and OTRS ticket systems
  • HuggingFace Local Plugin (otai-hf-local): Run ML classification models on your own infrastructure

How it works: Plugins are standard Python packages discovered via entry points. Install with uv add otai-otobo-znuny, reference in your config, and you're ready. Learn more in the Plugin System documentation.

Flexible Pipeline System ​

Build sophisticated automation workflows with sequential pipe execution:

  • Simple Pipes: Fetch tickets, classify content, update fields, add notes
  • Expression Pipes: Dynamic conditional logic with Jinja2 templates
  • Composite Pipes: Nest pipelines for multi-stage orchestration

Each pipe receives context from previous steps, executes its task, and passes results forward. Read the complete guide in Pipe System.

Dynamic Configuration with Template Rendering ​

Configure everything using YAML + Jinja2 for maximum flexibility:

  • Reference environment variables: {{ get_env('API_KEY') }}
  • Access pipe results: {{ get_pipe_result('fetch', 'tickets') }}
  • Conditional parameters based on runtime state
  • Type-safe configuration schemas

Services are defined once and reused across multiple pipes via dependency injection. Explore Configuration & Template Rendering for details.

Easy Installation ​

The easiest way to setup Open Ticket AI on your server is using Docker Compose:

1. Create compose.yml:

yaml
services:
    open-ticket-ai:
        image: openticketai/engine:latest
        restart: "unless-stopped"
        environment:
            OTAI_TS_PASSWORD: "${OTAI_TS_PASSWORD}"
        volumes:
            - ./config.yml:/app/config.yml:ro

2. Create your config.yml (see Configuration Guide)

3. Start the service:

bash
docker compose up -d

4. You also need to setup the Ticketsystem

Alternative: Install with pip/uv ​

For local development or custom deployments:

bash
# Install uv package manager
curl -LsSf https://astral.sh/uv/install.sh | sh

# Install Open Ticket AI with all plugins
uv pip install open-ticket-ai[all]

# Or install plugins individually
uv pip install open-ticket-ai
uv pip install otai-otobo-znuny otai-hf-local
bash
# Install Open Ticket AI with all plugins
pip install open-ticket-ai[all]

# Or install plugins individually
pip install open-ticket-ai
pip install otai-otobo-znuny otai-hf-local

See the full Installation Guide for system requirements and deployment options.

For Plugin Developers: Build and Monetize ​

Open Ticket AI 1.4 empowers developers to create and sell commercial plugins with complete licensing freedom. There's no marketplace yet, but the foundation is ready.

Plugin Development Freedom ​

  • No licensing restrictions: Choose your own license model
  • Sell commercial plugins: Monetize your extensions however you like
  • Full documentation: Complete guide at Plugin Development
  • Community visibility: Your plugin can be listed on our Plugins page

Future Marketplace ​

While there's no official marketplace today, we're building toward one:

  • Plugin listings: Already available on the documentation site
  • Discovery page: Coming soon with search, categories, and ratings
  • Community showcase: Highlight popular and trending plugins

Start building now, and your plugin will be ready when the marketplace launches!

Technical Highlights ​

  • Python 3.14: Modern type hints, performance improvements
  • Dependency Injection: Clean architecture with Injector framework
  • Entry Point Discovery: Standard Python packaging for plugin loading
  • API Compatibility Validation: Plugins and core versions checked at runtime
  • Comprehensive Testing: Full test coverage with pytest

Open Ticket AI 1.4 is production-ready, extensible, and built for the future. Install it today, automate your workflows, and join the growing plugin ecosystem!