Zammad

What Zammad does

Zammad is a web-based help desk that collects requests from many channels into a single ticket system. Email accounts (IMAP, POP3, Google and Microsoft 365), a chat widget for your website, a web form, SMS, Telegram, WhatsApp and Facebook all create or update tickets, and telephone systems connect through CTI integrations for Placetel, sipgate or a generic interface. Agents work with groups, roles, organizations, tags, overviews and checklists.

The automation layer is detailed. Triggers react to ticket changes, schedulers run time-based jobs, macros apply several changes in one click, and core workflows control which fields appear and when. SLAs run against business-hours calendars. A knowledge base can be public for customers or internal for agents, and time accounting records the effort spent per ticket.

Migration and AI

Zammad has official migrators from Zendesk, Freshdesk, Kayako and OTRS. They import tickets and articles, groups, organizations, agents and customers. A migration only works into a new, empty instance, from one source per instance, and it is all or nothing, so test it first on a separate installation.

The AI features (ticket summaries, a writing assistant, a knowledge base assistant and AI agents) are optional. They do nothing until you configure a provider: Zammad’s own service, OpenAI, Anthropic, Azure AI, Mistral AI, a local Ollama model or another OpenAI-compatible endpoint. One switch turns off all AI calls.

Self-hosting Zammad

Zammad is a Ruby on Rails application with a Vue and TypeScript front end. It can be installed from DEB or RPM packages for current Debian, Ubuntu, RHEL-compatible and openSUSE/SLES releases, with Docker Compose, on Kubernetes with a Helm chart, or from source. It needs PostgreSQL 15 or later, Redis 7 or later and Nginx or Apache as a reverse proxy. Elasticsearch is optional but strongly recommended, because search, reporting and attachment search depend on it.

The hardware requirements are higher than for most help desks. The documentation gives a minimum of 2 CPU cores and 6 GB of RAM, plus 4 GB more if Elasticsearch runs on the same server. Its example for up to 40 agents uses 6 cores and 6 GB of RAM, plus 6 GB for Elasticsearch.

The whole code base is AGPL-3.0. The README says the code is owned by the Zammad Foundation, independently of commercial providers such as Zammad GmbH. The repository has no proprietary directories.

Hosted version

Zammad GmbH runs a hosted service in German data centers with a 30-day free trial. No free plan is listed. The company also sells support contracts for self-hosted installations.

Limitations

Zammad needs more memory than most help desks here, mostly because of Elasticsearch. Migrations must go into a fresh instance, which rules out merging data from two old systems. The interface and admin settings are extensive, so setup takes some planning.

Who it suits

Zammad suits support teams that want a single open source help desk for email, chat, phone and messaging, especially those moving from Zendesk, Freshdesk or OTRS. It fits organizations that can run a moderately sized server or prefer a hosted service in Germany, and that want every feature under the same open source license.