You upload the recording of a talk, interview, meeting or podcast. The platform transcribes it, writes a new text in your brand's style and hands it to you in the editor to refine and publish — to LinkedIn, a blog or the brand profile — or share with your team. This page details every feature, with the formats, limits and languages the platform actually accepts.
We describe only what is live today. Where a limit exists, it is written down; where the platform does not do something, we say that too.
Each block describes what the feature does in practice, with its limits spelled out, instead of a generic promise.
The whole file goes to transcription: you never have to cut a two-hour talk into pieces before uploading it. The transcript comes back organized, separating who speaks and timing every passage, in the language you picked at upload.
The text is not a summary of the transcript: the AI writes a new article out of it, with editorial craft and in the language you choose at creation. Generation runs in the background, so you keep working and get told when the article is ready.
A style holds the instruction that governs tone and structure, plus an optional target length in characters. Set it once and every article follows it, and at each generation you can still add one-off instructions meant only for that particular request.
The persona describes who the text is for: role, context, familiarity with the subject, what they want to take away and what makes them stop reading. The AI uses this to keep the explanation level, choice of examples and cut right.
There are two layers. The modules written by professional writers ship with the platform and guide the writing craft. Your own base is optional: they are your operation's documents, whose text is extracted and used as context for that article.
One article can take several recordings as sources, and also articles you have already written on the platform, as long as everything sits in the same folder. The AI synthesizes them into a new text with its own thread, rather than splicing excerpts.
The article lands finished in the editor, and from there it is yours: type by hand, ask the AI to rewrite passages, or generate the article again with different instructions. Then share the text with your team or your client, per article or per whole folder.
Approval is the content's exit gate. Before anything goes out, the text is explicitly approved; after that it can be sent to WordPress or LinkedIn, or shared through the brand's public profile and secure links with the right readers.
Access is granted by invitation and per folder, not by a blanket permission. Whoever uploads audio never sees the editing, and the client receives only the finished text.
Configures the organization: integrations, policy, quotas and folder access.
Creates and edits articles, registers styles and knowledge and decides who joins each folder.
Writes and edits inside the writer's workspace, without inviting anyone or managing access.
Uploads audio to the folders they were invited to and sees nothing beyond their own uploads.
Uploads audio and follows the folder's finished articles, without editing anybody's text.
Reads only the finished articles that were shared with them, with no audio and no transcript.
Audios and transcripts stay restricted to the writer and their team. Uploader, partner and client only access the approved final content.
The questions that come up before starting, answered with the platform's real limits.
Access is by invitation, starts with a diagnostic conversation and continues with an editorial curator beside you through the first weeks.
Tell us what you produce today and our team reviews your operation to schedule the conversation.