Dashboards

A dashboard is a live, filterable overview of everything said in one or more discussions. Each substantive statement becomes a contribution card, and the AI can group, tag, and rate those cards so you spot patterns without reading transcripts.

What you see

• Cards — every contribution as its own card, with a short summary, the speaker and the source discussion.
• Topic view — contributions grouped into main topics and sub-topics once you run Extract themes.
• Sentiment — after you run Analyze sentiment, each card is rated Positive, Neutral, or Negative. Read it as a rough signal: criticism phrased as a hope can come out positive.
• Tags — define your own categories under Tag categories, each with a name and a one-line Category description (for example "Costs" or "Participation"), then run Tag in the AI Actions row to sort contributions into them. The AI reads each statement on its own and gives it every tag that fits; each card then shows its tags. If you have changed the categories, Tag asks whether to apply Only new categories or Re-tag all categories.

Two kinds of tags work well together. Thematic tags say what a statement is about, in your own terms; the themes above are the structure the AI finds. Functional tags say what kind of statement it is, for example proposals, barriers, values or narratives.

Use Filters to narrow the cards by discussion, speaker, theme, sentiment, or tag — and Clear filters to start over. The count shows how many statements match, so you read exactly those — for example every Negative card, or every card with one tag.

How it updates

New recordings and uploads flow in on their own: fresh contribution cards appear automatically as soon as their analysis finishes. Themes, tags, and sentiment behave differently — they only change when you rerun the matching action in the AI Actions row: Extract themes for topics, Tag for tags, and Analyze sentiment for ratings. Until then, new cards that do not fit the saved structure collect under Awaiting Theme Assignment; rerun Extract themes to sort them in.

Run the same actions from the chat

Everything in the AI Actions row also works as a plain request to the Co-Pilot: ask it to extract the themes, tag the contributions, or analyze sentiment, and it runs the same actions with the same results on the dashboard. You can even create the dashboard itself in the chat — see Ask the Co-Pilot to create a document or dashboard and Ask the Co-Pilot instead of clicking.

Check and correct themes and tags

• Check a tag before you rely on it: filter by it and read a sample of the cards.
• Remove a wrong tag on the card itself, or add a missing one with Add tag.
• Move a card to another theme under Theme assignment.

Write a report from the dashboard

Ask in the chat where you created the dashboard, and say "dashboard" in your request, for example "Write a report from the dashboard, one section per topic." Your topics become the headings. Tell the Co-Pilot how to use tags and sentiment, for example to order the points or to describe the mood of a section. A summary written without the dashboard chooses its own structure. Open the report from its card in the chat, and click Focus mode (fullscreen) to read it on the whole screen.

Open the original contribution

Click any contribution card. The discussion transcript scrolls to the passage behind that card and highlights it, so you can always check the exact words — who said them, and in what context.

Related articles

• Plans
• Share documents and dashboards by link