DAIDalusGet Early Access

Private beta · 2026

Three workflows. One copilot. You decide what ships.

DAIDalus is built around three things people actually do, not a feature taxonomy. Each workflow is short, opinionated, and ends with your approval — never the AI’s autonomy.

Workflow 1 · Quick-cut

Cut a vlog by deleting words, not scrubbing a timeline.

Drop a 5-minute talking-head clip in. Whisper transcribes it in about 20 seconds. You select the bad takes, the rambling, the false starts — and press Backspace. Each deletion strikes through grey; the preview updates instantly. No timeline scrubbing, no nudging clip edges, no playhead math.

Workflow 2 · Copilot cleanup

Filler words, dead air, and ragged cuts — handled by approval, not autonomy.

Type one instruction: “Clean up filler words, trim silences over half a second, and add smooth transitions at each cut.” The copilot returns proposals — delete-range mutations for fillers and silence (shown as highlights), insert-transition mutations at each cut point (shown as small badges between clips). You scrub the suggestions, toggle individual proposals on or off, swap a transition style for another preset. Nothing applies until you approve.

Workflow 3 · Animated captions

Instagram-style bouncy captions, in one prompt.

“Add animated captions, Instagram style.” DAIDalus generates word-by-word captions synced to your existing transcript, using the bundled Remotion animated-captions engine. Switch between a small set of preset styles in the inspector. No separate captioning tool, no manual timing, no SRT files.

All three ship in private beta.

First invites land soon.

Nothing ships you didn’t approve.

Join waitlist