After your narration transcript is processed, Velo takes you to the VeloTwin Setup page. This is where you configure the AI agent - picking the presenter voice, setting the tone and style, and choosing how long the final Velo should be.
Selecting a VeloTwin
The Select Your VeloTwin section shows your available AI presenters as a grid of cards. Each card shows a thumbnail and the presenter’s name with their accent (for example, “Drew - American” or “Viraj - Indian”).
Click the play icon on any card to hear a voice preview. Click a card to select that presenter.
Once selected, the grid collapses into a summary card showing your chosen VeloTwin, with a Change VeloTwin link if you want to switch.
Your VeloTwin cannot be changed after Velo generation. Choose carefully - this selection is locked once you click “Go, Make my Velo.”
Public VeloTwins vs. My VeloTwins
The Public VeloTwins tab shows Velo’s built-in library of AI presenters. The My VeloTwins tab shows any custom VeloTwins you’ve created. Learn how to create one in the Creating VeloTwin section.
Additional instructions
The Additional Instructions field lets you give the agent specific guidance on how to write the script. Use it to shape the tone, audience focus, or messaging priorities. For example:
- “Keep it conversational and jargon-free - this is for a non-technical audience.”
- “Focus on the key workflow, not the settings or admin pages.”
- “Highlight how this saves time for marketing teams.”
This field is optional, but adding a few clear sentences here meaningfully improves script quality.
Default style
The Default Style dropdown sets the overall presentation style. Options include styles like Default, Product Manager, and others - choose the one that best fits how you want your Velo to feel.
Duration
The Duration dropdown sets the target length for the generated Velo. Choose based on your intended use case - shorter for quick product teasers, longer for detailed walkthroughs.
Generating the script
Once you’ve set everything up, click Generate Script. Velo’s AI agent will analyse your narration transcript and the URL content to write a scene-by-scene script.
You’ll be taken to the Script Review page when the script is ready.