Skip to main content
Help Center

Introduction to AI features

What teamspace does for you with AI, why the AI is always your own – and how that differs from connecting your own assistant via MCP.

At certain points, teamspace can call on an AI to help: condensing a long ticket, drawing the key insights out of a report. These features are built in. You don’t need to connect anything, install anything or learn anything – an administrator enables them, and after that they are simply there, right where you need them.

Two ways to use AI with teamspace

In everyday talk, both are called “AI in teamspace”, but they mean different things. The distinction is worth making, because it determines who you need to ask when something is missing:

AI features (this topic)MCP
What happens?teamspace uses an AI internallyYour assistant accesses teamspace from outside
Where do you work?In teamspace itselfIn Claude, ChatGPT, Cursor …
What do you do for it?Nothing – the feature is thereConnect once and authorise via OAuth
Who sets it up?An admin connects an AI and enables featuresAn admin defines the access, you connect your client
ExampleCondense a long ticket history with one click”Show me all open bug tickets by status”

So if you connect your ChatGPT account to teamspace via MCP, that is not using AI in teamspace – it is your assistant using teamspace as a tool. Conversely, condensing a ticket needs no client and no MCP: it runs in teamspace, with the AI your company has connected.

The two don’t rule each other out. How the other way works is described in Introduction to MCP.

What works today

FeatureWhereWhat it does
Condense a ticket historyIn the ticketCondenses a long history into the matter of the ticket.
Analyse a reportIn the report, next to the Excel exportReads the report and names the notable points.

More is listed in Available AI features – the list keeps growing.

Three decisions that explain everything

When it comes to AI features, teamspace is deliberately more restrained than others in several places. Once you know these three points, the rest follows on its own.

It is always your own AI. teamspace brings no AI provider along where your data ends up. Every tenant enters its own endpoint. That keeps the question “where does our data go?” one that your company answers – not us. Why this makes all the difference for data protection is explained in Your AI, your data.

The AI supports, it doesn’t decide. You won’t find a feature in teamspace that writes an invoice on its own or closes off an operation. AI suggests, summarises, points things out – the decision stays with you. This isn’t caution for the time being, it is a design decision: Why teamspace relies on rules for AI.

Where it has to be exact, rules do the calculating. Invoices, billing, figures: for these, teamspace has a rule-based system that delivers 100% correct results – not 99.9%. There, AI is used at most as a reviewer, never as a calculator.

What you have to do

As a user: nothing. If a feature is enabled, it appears in its place – in the ticket, in the report. If it isn’t there, your administrator hasn’t enabled it (yet), or no AI is connected.

As an administrator: connect the AI once, then enable features. See Connect an AI.

Availability: The AI features are included in all editions. From September 2026 they are enabled for all tenants; prospective and beta customers can test them on request already. More features will be added in the meantime.