Zammad AI¶
Note
Zammad’s AI features are currently limited to SaaS customers.
This page may change regularly.
Consider this page in Beta state.
Introduction¶
Zammad offers AI powered features to support your agents! The configuration is pretty easy so the entry barrier is low.
Read on how to configure it and which features are currently supported.
Tip
🚀✨ What’s the easiest way to get AI features running? 🚀✨
Subscribe to a plan with Zammad AI, go to AI > Provider in the Zammad settings and choose Zammad AI. Enter the token you received and you can start using the features by simply enabling it.
Provider Configuration¶
To use AI feature in Zammad, you have to configure an AI provider which handles the requests. You can configure it in Zammad’s admin settings under AI > Provider.
The following providers are supported:
Note
The AI service you have subscribed to is not available? This is just the first step. We are going to add more providers soon! If you want to support us or need the provider as soon as possible, just let our sales department know!
Ollama¶
Ollama is an AI tool that allows you to set up your own AI server on a machine you control. Be aware that this only makes sense if you have a powerful GPU in your system!
Using Ollama on an own server makes sure that your data is not transferred to a third party.
When choosing Ollama as your AI provider, you just have to provide the IP/URL
of your instance with appended port number. After that, make sure to click
the Save button and read on in the Features section. If
you don’t have an Ollama running already, their
deployment guide
is a good starting point.
Depending on the feature you want to use, make sure the right model is available. If you just installed Ollama, you can pull and run a model by executing:
ollama run llama3.2
You can find the available models in Ollama’s model library and useful information and additional links in the Ollama Github repository.
OpenAI¶
You probably already heard about OpenAI, haven’t
you? To use their service as AI backend, choose them as AI provider and add
your API token. After that, make sure to click the Save button and read on
in the Features section.
Zammad AI¶
Tip
Don’t want to deal with all this stuff or just want to have a privacy respecting AI service? No problem! Get in touch with us to get your token and you are ready to start.
The usage of Zammad AI as a provider has some benefits:
You don’t have to care about the setup and configuration compared to your own setup.
Your requests with sensitive information aren’t used for training.
If something isn’t working as expected, you just have one place where you get support.
To configure it, simply choose Zammad AI as your provider and add your
token. After that, make sure to click the Save button and read on in
the Features section.
Features¶
Below you find the supported AI powered features of Zammad. Make sure to configure a provider before activating a feature. Otherwise, a warning message will tell you to do so.
Note
Currently, only one feature is available so far. This feature can be seen as the foundation for other useful features. Expect more to come soon!
Ticket Summary¶
Activating this feature generates summaries of tickets. This can be a huge time saver when dealing with large tickets and/or many hand-overs between agents.
By default, Customer Intent and Conversation Summary are enabled and always active. You can choose to activate or deactivate Open Questions and Suggested Next Steps, depending on what fits your needs.
Enabling the ticket summary feature triggers a request to the configured AI service when an agent opens a ticket. Agents then see a banner below the ticket articles which tells them that a summary is available. They can now simply click the included button to jump to the AI Summary sidebar tab. The banner can be hidden by agents. In case an agent wants to reactivate it, this can be done in the user’s profile settings under Appearance.
Hint
If you use an own Ollama server as provider, make sure to have the
model llama3.2 available for this feature.