Mosaic is a platform that enables users to automate video editing workflows through a node-based canvas interface. It allows creators to set up scalable video editing processes that can handle everything from rough cuts to motion graphics, providing an alternative to traditional timeline-based editors.

The platform features a node-based canvas where each node represents a video editing operation configurable with natural language prompts. Users can branch workflows to run edits in parallel, creating multiple variants from the same footage for A/B testing. The system includes inline timeline editing capabilities and supports exporting timelines back to traditional editing tools like Premiere Pro, Final Cut Pro, and DaVinci Resolve.

Mosaic uses multimodal AI to build visual understanding and intelligence into its agentic workflows. This includes computer vision and video processing techniques like saliency analysis, audio analysis, and object significance detection to guide editing decisions. The platform can understand video concepts, emotions, actions, spoken word, light levels, and shot types.

Benefits include accelerating video editing from hours to seconds, enabling A/B testing of different content variations, and providing scalable content creation workflows. Use cases include removing bad takes, creating script-based cuts, repurposing long-form content into clips and reels, creating sizzle reels, generating assembly edits, optimizing content for social media platforms, dubbing with voice cloning, and creating motion graphics.

The platform targets content creators, video editors, and marketers who need to produce video content at scale. It integrates with traditional editing software through XML export and can be triggered programmatically via API or event-based triggers.