OTAI Runtime
On-prem connectors for OTOBO, Znuny/OTRS, and Zammad
OTAI Runtime
OTAI Runtime v3 is the on-prem workflow runner. It polls your ticket system, queries a local OpenAI-compatible LLM, and writes results back to OTOBO, Znuny/OTRS, or Zammad. Data never leaves your infrastructure.
Supported ticket systems
| System | Connector | Setup guide |
|---|---|---|
| OTOBO / Znuny | Generic Interface (REST) | OTOBO / Znuny setup |
| Zammad | Personal access token | Ticket system integration |
Set OTAI_TICKET_SYSTEM_TYPE to otobo, znuny, or zammad and set the matching variables. See Ticket system integration for details.
What v3 runs
Three workflows configured via OTAI Studio or REST API:
| Workflow | Purpose |
|---|---|
queue_classification | Route tickets to the correct queues or groups |
priority_classification | Set ticket priority from content |
ticket_summary | Write a summary to a custom field |
Key principles
- On-prem only — Zero cloud calls during inference.
- Env-driven — Credentials configured via environment variables.
- Local LLM — Point
OTAI_LLM_BASE_URLto Ollama, vLLM, or any OpenAI-compatible server.
Next steps
- Follow OTOBO / Znuny setup or configure Zammad variables.
- Read Ticket system integration for env vars and API details.