Edit by talking, not by dragging clips

First your audio is transcribed with word-level timestamps. The transcript is then split into moments: a concrete noun, an action, a statistic you mention, a metaphor. Each moment becomes a search query.

Queries are run against licensed stock libraries and ranked by how well the clip's description matches the sentence around it, not just a single keyword. "Our funnel was leaking" should not return a plumber unless you are talking about plumbing.

Every suggestion keeps a link to the exact words that triggered it, so you can always see why a clip was chosen.

Placement is decided from the transcript and the audio. BrollBot looks for sentence boundaries, pauses and emphasis, and avoids covering the speaker during short, punchy lines where the face carries the moment.

Cutaways have a minimum and maximum length so the edit keeps rhythm, and a talking-head beat is kept between B-roll blocks so the viewer does not lose the speaker.

Overlays such as lower-thirds and key-phrase callouts are placed from the same transcript analysis. You can change the density from sparse to busy before the first pass.

The AI makes a first pass; you make the final call. Click any B-roll clip on the timeline and a drawer opens with alternative matches for that exact moment.

Drag the clip to move it, drag its ends to trim, or remove it to return to the talking head. Changes are non-destructive, so your original recording is never altered.

Because every clip is tied to a phrase in the transcript, editing the transcript text re-runs matching only for the affected moment.

Captions use the same word-level timestamps as the matching step, so highlights land on the word being spoken.

Pick a style, set the font, color and position, and fix any mis-heard word in the transcript. Fixes update the captions and the matching together.

Captions are safe-area aware for each export ratio, so a 9:16 export does not hide text behind platform UI.

Upload a logo, choose your fonts and set brand colors. Captions, overlays and lower-thirds pick them up automatically.

Agencies can keep a separate kit per client and switch between them per project.

Kits apply at render time, so changing a color updates the next export without redoing the edit.

Choose the ratio when you export. The talking head is re-framed for each ratio and captions are repositioned to stay inside safe areas.

Pick a resolution that suits the destination. Exporting the same edit in several ratios costs credits for each finished video, as defined on the pricing page.

Files are rendered on our servers and available to download for the retention period.