A presenter inside the complete video
Martius combines an AI avatar presenter with the material that shows your product: real footage, a recorded workflow or a reviewed diagram. Use the person to introduce a launch, guide a demonstration or explain a difficult relationship.
Choose an existing presenter, an authorized custom portrait or user-provided presenter footage through the supported source paths. Keep that source identifiable; a custom character is not automatically a digital replica of a real person.
Choose the person, voice and approved script
Select the presenter and voice, confirm permission for their use and supply the approved lines. Store pronunciation guidance for product names and technical terms alongside the brand context.
Reuse the selected person and voice in later passages or project versions. Change the script and regenerate the affected presenter material without turning the presenter into an isolated asset disconnected from the film.
- Existing, custom-portrait or supplied-video source
- Authorized person and selected voice
- Approved script and pronunciation guidance
- Reusable presenter reference and applicable scope
Give the product enough room
Use a standalone presenter for an introduction or closing, a corner view while a demonstration remains prominent, or a split view when person and visual need equal space. Adjust when the presenter appears and how much of the frame they occupy.
Keep prices, labels, interface controls and captions readable. Remove the presenter from a shot when the product needs the full frame; a familiar face should help the explanation rather than obscure its evidence.
Revise the message and review the voice
When product wording changes, revise the approved script and inspect the related presenter passage, narration, captions and scene timing. Compare with the parent version so the explanation remains connected to the current product.
Watch and listen to the complete result. Check pronunciation, lip sync, identity, transitions, sound quality and supported language treatment. Permission to use a face or voice does not establish the truth of what the presenter says.
Keep source provenance and data scope explicit
Original portraits, presenter footage, source audio and editable projects remain on your device. Model requests and derived data used by a particular generation path have a separate documented service scope; local source files do not imply fully offline AI.
The moving presenter example on this page comes from the team’s existing X-Pilot collection. Inspect its picture and sound directly. Its source is identified separately from the Martius custom-presenter workflow described here.