Eddie AI has released Logging v2, an update aimed at giving editors more control over how the system analyzes and organizes source material.
The main addition is topic steering. Instead of asking the AI to decide unaided what matters in a clip, editors can now give it prompts for specific themes, characters, locations, product names, objects or visual moments. CineD reports that the feature supports up to five topics or categories per clip in Eddie’s Docs/Stringouts mode.
That is a useful shift for documentary, interview, branded-content and unscripted workflows, where logging is not just transcription. Editors often need to find the right argument, recurring subject, reaction shot, location reference or cutaway buried inside hours of material. A steerable logging pass should, in theory, produce metadata and stringouts that are closer to the story the editor is trying to build.
Logging v2 also adds support for backgrounder documents. Editors can attach a Google Doc, PDF or Word file during import so Eddie has more context before it proposes story suggestions. In practice, that could mean feeding the system a treatment, interview plan, research document, shot list or client brief before it starts organizing the footage.
The update also extends still-image support for B-roll workflows. Eddie’s help documentation says stills can be imported as B-roll in Docs/Stringouts mode, where they are converted into short video clips for use alongside other B-roll material. The company has also increased its Pro+ tier to support up to 20 hours of source material in Docs/Stringouts projects.
The important point is not that the AI is “editing” on its own. It is that logging tools are starting to move from passive transcription toward brief-aware media organization. For editors, the value will depend on whether the prompts and background documents produce cleaner bins, better stringouts and fewer tedious search passes, rather than just another layer of AI confidence draped over messy footage.