Datafisher vs Scribe

Scribe's recent work splits cleanly in two: making the Optimize agent answer from an organization's own context, and broadening what can become a Scribe in the first place. Admin-set business context now feeds every agent chat, and video uploads join browser capture as a source for step-by-step documentation. Around those, the team is filling in workspace basics — departments, cross-team sharing, and now a sortable list view for the library.

The capture engine is being decoupled from the browser recorder: if any video can become a Scribe, the product's input surface stops being tied to a live session. At the same time the agent is moving from a generic assistant to one that reasons over an org's tools and goals, which is what makes it defensible against a general-purpose chatbot pointed at the same documents. Library and org management work is following the customer size that grounding attracts.

Expect the next moves to push the Optimize agent further into retrieval over the library itself, and to extend the video-to-doc path with more source types or bulk import. Library management is likely to keep gaining filtering and ownership controls as document counts grow.