Comcast Technology Solutions has announced six AI-powered media applications for video verticalisation, localisation, smart chapters, metadata enrichment, quality control and intelligent clip generation. The functions build on its VideoAI framework and will be offered across Comcast Media360, Comcast Sports360 and standalone Cloud Video Platform deployments.
The applications connect to Comcast's Cloud Media Processing engine, where an AI function can be inserted as a pre-processor, content-manipulation step or post-processor. That design matters operationally because a customer can add one function to an established channel or library workflow without rebuilding the complete processing chain.
The verticalisation tool identifies an action centre and creates cuts or pans for portrait outputs. Localisation separates and translates dialogue while retaining background sound, then adds dubbing, voice matching and captions. The other applications divide programmes into chapters, enrich descriptive data, flag quality problems and prepare clips for editorial or promotional use.
Comcast says the suite is based on implementations used by large media companies, but it has not published model-level accuracy, supported languages, human-approval controls or customer-specific performance data. Commercial availability will be phased through 2026 and 2027, so the announcement describes a product programme rather than one simultaneous general release.
