Creator Studio

What you can create

Podcasts — the AI generates a transcript. Edit it manually or with the AI, choose a voice for each speaker, then generate the audio.

Power reads — the AI generates a one-page power read. Edit the text and format it using the same options available when writing one by hand.

Videos — a slide-based video generated from your documents. You can edit the slides, the transcript and the speaker voices, and adjust slide background and text color.

Scripted videos — the AI writes a script you record with the Teleprompter. Available when scripted videos are switched on for your organization.

Source materials

Each piece starts from the files you add under Source Materials. There are two ways to add them.

Uploaded Files — drag files in, or choose them from your computer. These files are used only for this piece. They are removed if you leave the page before generating, and they are not added to the Knowledge Base.

Knowledge Base — choose Browse Knowledge Base to pick files already in the initiative's library. Only files that have finished processing can be picked. See Knowledge Base.

You can also start from the Knowledge Base itself: select files there and choose Create AI Content. The creation screen opens with those files already added.

Supported files are PDF, Word, PowerPoint, OpenDocument, video and audio. Documents can be up to 200 MB, audio files up to 1 GB and videos up to 3 GB. A file that is too large or of an unsupported type is marked Not uploaded with the reason. Remove it and upload a different file. For best results, convert PowerPoint files to PDF before uploading.

If you don't have access to the initiative's Knowledge Base, files that someone else added from it are hidden when you open the content, and the source materials list shows "You don't have access to some files."

Context capacity

The Context capacity meter shows how much of the AI's limit your files use. When it is full, or a file doesn't fit, remove a file to make room.

The three steps

1. Add source materials and a prompt. Add your files and write a prompt to guide the AI. The prompt shapes tone, style and audience. You can also choose speaker personas here.

The first draft takes a few minutes. A checklist shows its progress, and you can leave the page and come back. The AI assistant becomes available once the first draft is ready. If generation doesn't finish, no content is saved. Open the AI assistant to try again.

2. Edit the transcript and voices. Change anything the AI produced — rewrite a sentence, add context, adjust formatting. Use the AI panel to revise the prompt if the whole output needs to move in a different direction, and choose the voices.

While the assistant is working, choose Stop to end its response. On a video, the assistant can edit the slides for you. You can browse slides while it works, and stop the response to edit them yourself.

You can use more than one prompt on the same source materials. Refining the prompt and regenerating is a normal part of the process.

3. Publish. Add a cover image, complete the content fields and publish.

Saving your changes

Podcast and power read edits save automatically. The editor shows Saving, then Changes saved. If a save fails, choose Retry.

If you try to leave an editor with changes that haven't been saved, you are asked what to do with them. Moving on to the next step offers Save & Continue or Discard Changes. Leaving any other way offers Discard & Exit.

Once generated

Studio content cannot be taken back to the generation step to be regenerated. Edits after that point are made to the output, not to the source.

Studio content opens in Studio rather than the standard editing form. In the content table, filter by source to separate Studio content from manually created content.

Studio content is revised the same way as any other published content. See Revising published content.