Faster voice and caption generation
Many of our newsroom partners pointed out that the second step – generating voice and captions – sometimes took too long, with noticeable differences between runs.
We dug into this and optimized the process significantly.
The average generation time has been cut from about 60 seconds to around 30 seconds, which makes a real difference in everyday work.
This update has been very well received by our users and clearly improves the workflow comfort for editors.
Template architecture rebuilt
A less visible, but very important change: we’ve rebuilt the architecture for managing template styles.
With more clients onboard – each with their own custom template – the old setup was starting to show its limits.
We initially approached it in a typical startup way (MVP first, architecture later), which created some technical debt. This release clears that up.
The new, lightweight structure makes it easier and faster to onboard new titles and styles into the system.
Custom templates and voices
Many newsrooms that started testing Flipatic a few months ago are now staying on – and moving to their own unique templates.
They usually begin with a generic one, just to get started, and later we run short workshops together to define their personalized style.
We align it with their branding, website design, and visual identity, and often create custom voice models to match their tone and personality.
The difference in the final videos is clearly visible – both visually and in sound.
New titles joining Flipatic
In November, seven new newsrooms joined Flipatic – from Ireland, the Netherlands, Luxembourg, and Poland.
We’re excited to see the growing community of publishers using Flipatic across Europe.
Photo by RDNE Stock project.
