April 7, 2026

Mosaic Raises $3.8M to Build Video Editing Agents

What started as a side project to edit our own YouTube videos has quickly turned into something much bigger. Today, global agencies, platforms, and news networks rely on our AI video editing workflows to scale content production.

Fundraising Announcement

Mosaic has raised a $3.8M seed round to build video editing agents.

What started as a side project to edit our own YouTube videos has quickly turned into something much bigger.

Today, global agencies, platforms, and news networks rely on our AI video editing workflows to scale content production. I am especially excited to announce our partnerships with:

  1. TubeScience: Meta's largest ad creative partner, 8K videos a month, 100M views a day.
  2. News Corp: one of the world's largest media organizations, owning businesses like The Sun, TalkSport, Wall Street Journal, MarketWatch, HarperCollins, The Times, and more.

The funds from this round will be used to scale up our lean team in San Francisco and continue research & development on the frontier of multimodal AI and agentic video editing.

We are proud to be backed by top class Silicon Valley investors, like Y Combinator, Mayfield Fund, Elevation Capital, Pioneer Fund, Transpose Platform, TwentyTwo Ventures, Script Capital, Phosphor Capital, Amino Capital, Olive Technology Ventures, Fog Ventures, DCG, and angels from YC, OpenAI, Google DeepMind.

Our First Product: Canvas

We are living through a generational moment, perhaps one of the most consequential in the history of all of humanity.

The technologies and interfaces that will come out of this era will be so uniquely different from any past product experiences.

For builders, it becomes critically important to re-evaluate every problem, every solution, and every supposed constraint from first principles.

This is the thinking with which we approached our first product, the Canvas.

While every other AI video editor attempts to retrofit the traditional non-linear editor with a chat copilot, Canvas introduced a completely new interface for it.

Canvas is:

  1. Scalable: made for high-volume video workflows built around scalability and automation.

  2. Customizable: custom rules around brand styles, brand guidelines, and video formats.

  3. Extensible: custom nodes for specific workflows and niche use cases.

  4. Seamless: integrates with where you already work — source assets from central MAMs like Mimir and Iconik, export editable timelines back to NLEs like Premiere and DaVinci, and get updates in Slack throughout the whole process.

  5. Infrastructure: programmatically invoke video workflows through API or event-based triggers like Google Drive, Dropbox, AWS S3, and YouTube listeners.

    • We also now support agent-to-agent interaction (Agent Skills available for OpenClaw and Claude Code).

Today, global agencies, platforms, podcasters, and news networks rely on our AI video editing workflows to scale content production.

Canvas is made for video workflow automation.

Canvas is where you craft your Mosaic.

Mosaic Manifesto

The intersection of AI + video is one of the most interesting domains to be operating in.

In her article, Justine Moore (partner at a16z) does a great job describing the State of the Union for agentic video editing and the technological unlocks that enable it.

Video is increasingly the medium through which moments are captured and ideas are communicated. Yet every video — from conception to upload — still goes through the critical bottleneck that is video editing.

Different companies approach this problem in different ways.

Some leverage AI to solve small slices of traditional video editing problems, like searching through footage and clipping for content repurposing. While these are powerful and lucrative businesses, they represent smaller pieces of the larger puzzle. None yet provide a comprehensive and general-purpose solution to agentic video editing.

Others aim to remove the need to edit all together, focusing solely on generative media and orchestration. While the possibilities for creation and storytelling are unbounded here, models still hallucinate, are expensive, and hard to iterate with. That being said, there is still a lot of exploration and untapped potential here.

At Mosaic, we have the following convictions:

  1. Orchestration-first: UI is melting away. The orchestration layer is where the magic happens. For video editing, a really good orchestration architecture is one which can take a creative vision and transform it into a video. Once the orchestration layer is built, the interface for it can be the CLI, a simple chat input, a node-based canvas, voice dictation, or anything else.
  2. Editability is a must-have: Video is inherently visual and subjective. While compilation and logical checks can be performed on the output of a coding agent, there is no such objective checks to make sure a video agent did its job "correctly". Even trying to define what "correctly" means in this context has no clear answer. That's why the ability to edit and iterate on outputs quickly is essential for creative control.
  3. Generative Media: Future generations won't be averse to "AI slop," they'll grow up with it. AI media and videos (in the limit, realtime personalized generative media) certainly represent a large part of the future of content.
  4. Video Captures Human Experiences: Even with all the AI content in the world, humans will never stop watching other humans. And humans will never stop sitting in front of a camera to tell their own story. Even if it's a story that has been told a thousand times before by a thousand others. They'll tell it all the same.

Our solution to agentic video editing will be one which is built on these principles.

FundraisingSeed RoundAIVideo EditingAgents

MOSAIC