# Mosaic: Agentic video editing with a human finishing pass

Canonical: https://mudpie.ai/companies/mosaic/
Breadcrumb: [Home](https://mudpie.ai/) / [Companies](https://mudpie.ai/companies/) / [Mosaic: Agentic video editing with a human finishing pass](https://mudpie.ai/companies/mosaic/)
Author: Ali Abouelatta (https://mudpie.ai/authors/ali-abouelatta/)
Published: 2026-09-19
Updated: 2026-09-19
Research type: Company profile
Method: Company and accelerator sources checked 2026-09-19. Product claims are attributed to their sources; this is research, not a hands-on product trial.

Mosaic is an agentic video-editing canvas for turning raw footage into repeatable edits and variants. It fits a creator, media or marketing team that wants automation to handle the first large pass while keeping a familiar editor for the final taste and polish.

## What it does

[Mosaic’s current product description](https://mosaic.so/) presents a canvas where video-editing agents can run on autopilot, create multiple variants from the same footage and then hand the result to a timeline editor. [The YC profile](https://www.ycombinator.com/companies/mosaic-2) says the company is trying to encode the language of video and give that taste to agents.

The product history explains the interface choice. The founders say they first built a chat copilot, then found prompt-response UX too sequential for video. They moved to a node-based canvas where a team can create and run multimodal editing agents, then use chat and the editor for final touches. The launch describes the agent getting a project 80–90% of the way there; that is a company product claim, not an independent quality benchmark.

This is a better fit for a repeatable editing system than for a one-off cinematic project. A team can encode a format, run several cuts from the same raw material and reserve human time for the decisions that affect story, pacing and brand. The value depends on whether those decisions can be expressed as reusable nodes rather than remaining inside one editor’s intuition.

## Founder context and tradeoffs

The YC profile identifies Adish Jain as CEO, with engineering experience at Tesla, AWS and Berkeley, and Kyle Wade as CTO, with machine-learning research at UCSD and software engineering at Tesla. The founders say they met the problem while editing a video after work at Tesla. The context supports the product direction, not a guarantee of final-cut quality.

Pricing is not public. A buyer should ask what models and rendering costs are included, how assets and project state are stored, whether agents can be versioned and what export or collaboration controls exist. The canvas can improve throughput, but it also adds a new layer to learn and govern.

## Editorial take

I would shortlist Mosaic for a team publishing recurring video formats from a steady stream of footage. I would not choose it because “agentic” sounds faster. The proof is one real production format: can the canvas generate useful variants, preserve the team’s style and leave an editor with less repetitive work?

## Quick facts

| Field | Sourced detail |
| --- | --- |
| Product | Agentic video-editing canvas with timeline handoff |
| Buyers | Creators, media teams and marketing organizations |
| Public workflow | Node-based multimodal agents, variants and final editor polish |
| Pricing | Not publicly listed |
| Main gate | Repeatable formats, asset handling, model cost and human review |

## Sources checked

| Source | Checked |
| --- | --- |
| [YC profile](https://www.ycombinator.com/companies/mosaic-2) | 2026-09-19 |
| [Mosaic homepage](https://mosaic.so/) | 2026-09-19 |
| [Mosaic seed announcement](https://mosaic.so/blog/mosaic-seed-round-announcement) | 2026-09-19 |

## Cohort context

Mosaic is listed in Winter 2025. In our 2026-09-18 directory snapshot, 8 of 165 listed companies in that cohort have YC’s primary industry label Consumer (4.8%). This is a current-directory comparison, not an original intake count or a performance ranking. [Nine-cohort dataset](https://mudpie.ai/research/yc-cohorts-2026-09-19.json).

## Public website snapshot

Observed 2026-09-19T16:19:40.529Z in raw homepage HTML. This records visible metadata and advertised links, not agent execution or product quality.

| Signal | Homepage observation |
| --- | --- |
| Product description metadata | Observed |
| Canonical link | Observed |
| H1 or H2 heading | Not observed in this response |
| Typed structured data | Not observed in this response |
| Docs/developer link | Not observed in this response |
| Pricing link | Not observed in this response |
| llms.txt link | Not observed in this response |
| Markdown alternate | Not observed in this response |

[Public observations](https://mudpie.ai/research/yc-homepage-links-2026-09-19.json) · [Collection method](https://mudpie.ai/research/yc-homepage-methods/README.md). Missing links here do not establish that a capability or file is absent elsewhere.


## Author disclosure

I cofound Lazyweb and publish Mudpie. This is an owner-written publication, not an independent testing organization. Research notes distinguish observations, sourced reporting and editorial judgment.