Share ideas.
Take the next step together.

1. Open your shared board

Click the link you received from the event organiser. It opens your shared noticeboard. In the tool, this is called a “board”.

Only received a code? Open the network tool on your event platform. Enter the code in the “Event code” field and click the arrow beside it.

2. Write down your idea

A contribution is called an “input” in the tool. The “Input” section has three fields:

• Topic: What is it about?
• Challenge: Where do you need help?
• Idea / next step: What would you like to try?

Filling in one field is enough. You can also use all three. Keep it short and specific. Try to focus on one thing per contribution.

For example:

Topic: Use less electricity at work.

Challenge: We do not know which machines use the most electricity.

Idea: We are looking for another business to measure electricity use and share experiences with.

Under “Add your perspective (optional)”, you can add your industry or region. This helps others understand your contribution. You can also leave these details out.

Under “Pop-up in your city (optional)”, you can enter a city and country. Where would you like a pop-up implementation atelier? These are hands-on workshops where people work on practical ideas together. Your details help us plan suitable cities. You can fill in both fields or just one. We do not need an address.

Your city and country are visible with your contribution on the board and included in exports. If you do not want this, leave the fields blank.

3. Save your contribution

Click “Put input to work” at the bottom of the form. This means submitting and saving your contribution.

Wait for the message “Contribution saved.” You can then add another idea.

Important: Your contribution is not saved until you submit it. Wait for confirmation before closing the window.

4. Discover similar ideas

Click “Refresh results” to see new contributions.

The tool groups similar topics together. These groups are called “clusters”. Click “View inputs” to read the contributions in a group. A group's name does not fit? Tap “✎ Change title” there. Everyone then sees the new title.

Want to see everything? Choose “View all inputs”. You can also search for a word there, such as “energy” or “education”.

5. Talk in the cluster: transcribed and translated live

Would you rather talk about a topic group than type? Open the cluster with “View inputs”. In the “Live transcription” section you will find two QR codes. Scan one of them with your smartphone camera.

• “My device”: Choose your language, confirm consent and tap “Speak and listen”. Start speaking straight away: “My name is …” is optional. You can add or change your name and language later. You hear other contributions in your selected language; the cluster’s main language does not change this personal choice.
• Only want to listen? Under “My device”, choose your language and tap “Just listen”. Your microphone stays off. You also hear contributions originally spoken in your selected language.
• “Several on one phone”: Place the smartphone in the middle of your group, confirm consent and tap “Start microphone”. You can speak immediately. Up to four people can optionally add a named voice using “Add voice (optional)”. Speak clearly, one at a time; check the transcript’s speaker assignment and correct a wrong name using “✎”.

Was a name recognised incorrectly? Tap “✎” beside the name, correct it and tap “Save”. In “My device”, change your language under “Your language”.

For automatic playback, “Sound on” and “Autoplay on” must be enabled. Active switches have a dark background and a tick. With autoplay off, start contributions yourself. Enable “Headphones on” to speak and listen at the same time; without headphones, recording pauses during audio playback to avoid echo.

A green border marks the contribution currently playing. With autoplay on, the inner transcript list follows audio playback. If you scroll yourself, your position is preserved. Tap “Current audio output” to bring the playing contribution back to the top and resume following.

In “Transcript of this cluster”, contributions appear in their original language and, when different, also in the main language. Choose the cluster’s main language under “Main language” (default: German). It is independent of each smartphone’s listening language. New contributions and translations appear automatically.

Under “People and contributions”, use “Search by title” to find a displayed name or title. Under “Sort by”, choose “Title A–Z”, “Title Z–A”, “Content: 0 → 100 %” or “Content: 100 → 0 %”. Content share counts words in the transcript; it does not rate contribution quality or measure speaking time. Use ↺ to reset search and sorting.

The “People” tab shows “Speaker squares”: each square’s area represents its content share, and even 0 % remains visible as a small square. Tap a square or list entry to select a person. The “Word cloud” tab shows up to 40 frequent terms in the main language, excluding filler words. Larger text means more frequent mentions.

In both views, zoom using + and −, the mouse wheel or two fingers, and drag the background to pan. Use ↺ to reset the view. You can drag squares and words. Moving elements respond to each other; use “Motion on” or “Motion off” to pause or resume movement. While the live room is connected, the list, content shares and word cloud update automatically through the live connection; you do not need to reload the page.

After consent is confirmed, voice cloning is prepared automatically once suitable speech samples can be clearly assigned. The GPU is preferred for speech recognition and audio output. During startup or when support is unavailable, OpenAI can handle recognition and ElevenLabs can handle audio output. The current GPU voice-cloning model supports German, English, French, Spanish, Italian, Portuguese, Russian, Chinese, Japanese and Korean. The language menu therefore does not mean that every language already supports a GPU voice clone.

If a particular voice clone is blocked by the speech provider because verification is missing, a voice labelled as a standard voice can take over. Your selected listening language stays the same. A standard voice is not a voice clone. Other provider errors can still prevent audio output.

Use “Download as text” to take the transcript with you. “Microphone paused” means recording is paused. Tap “Finish” to end the session.

6. Find out what you can do together

Under “Cross-connections”, the tool suggests contributions that might fit together.

Click “Compare inputs” and read both contributions. Discuss: Do we have a similar problem? Can we learn from each other? What could we try together?

The suggestions are there to help you think. You decide whether they make sense. The tool does not contact anyone or arrange conversations.

A useful next step is small and specific: Arrange a short conversation or a shared experiment, for example. Agree who will take care of it and when you will follow up.

7. Take the results with you

Want to keep working with the contributions after the event? Choose “Download CSV” in the contribution list. You will receive a file you can open in Excel, for example.

If you selected a topic group or search results first, “Selection as CSV” downloads only those contributions.

Each contribution stays on the board for 30 days after it is saved. Download the results in time if you want to keep them longer.

Organising your own event?

Here is how to provide a shared board:

1. Open the network tool and click “Start a new board”.
2. Open “Share board”.
3. Click “Copy link” and share the link with your participants.

If copying does not work, select the displayed link and copy it yourself. Ask everyone to use the same link so that all contributions end up in one place.

Tip for live transcription: Show a cluster page with its QR codes on a large screen in the room. The QR codes are valid for 30 days. Reload the page afterwards to get new codes.

What you should know about your details

Anyone with the board link or event code can read the contributions. Please do not include confidential information.

You can leave a way to reach you under “Add contact details (optional)”. You must agree to these details being stored. Details in this contact field are not displayed publicly and are not included in downloaded files.

Live transcription: The microphone starts only after you consent. Text remains visible in the cluster for 30 days. Clearly assigned speech recordings are stored encrypted for 30 days and may be used only to train speech recognition for this room. Introductory samples and suitable conversation contributions for a temporary AI voice clone are cached for up to 2 hours. The application requests deletion of the clone in the next cleanup run after 24 hours have elapsed.

Anyone with the board link can read the transcript. It also appears in the platform’s internal WirkFabrik channel. For transcription, translation and reading aloud we use SpeakSphere with services from SupraTix.

If something goes wrong

• The code does not work? Use the full event link or ask the event organiser.
• The form shows an error? Read the message beside the field and correct your entry.
• Not sure whether your contribution was saved? Check the board before submitting it again.
• Topic groups are not showing? Try again later. You can often still read the contributions using “View all inputs”.
• The microphone does not start? Allow your browser to access the microphone and reload the page. Live transcription works in all current browsers on iPhone, iPad, Android, Windows, macOS and Linux. If the link opens in an app's built-in browser, for example Instagram, open it in Safari, Chrome, Firefox or Edge.
• The QR code is no longer valid? Scan the current code on the screen in the room.
• Can’t hear anything? Check the selected listening language, media volume, “Sound on” and, where shown, “Autoplay on”. Keep the page in the foreground and the screen on. If no new contributions arrive, end the session, reload the page and start again. Text alone does not confirm successful audio output. Recognition, translation, audio generation or a playback queue can cause delays; a fixed latency is not guaranteed.