Palette raises €3M pre-seed to develop an OS for AI-native teams

Palette, a Copenhagen-based AI software startup founded in 2025, has raised €3 million in pre-seed funding to push its platform beyond early adopters and into more parts of the business. The round was led by Ugly Duckling Ventures, with Emblem, Acadian Ventures and a group of angel investors also backing it.

The company is building what it calls an operating system for AI-native teams. The pitch is simple enough: as companies bring more AI tools into sales, operations and product work, they should not have to rebuild their internal context every time a new model shows up. Palette’s idea is to keep the company’s workflows and knowledge separate from the model underneath them, so the model can change while the organisation’s own setup stays intact.

That separation matters because Palette is betting the next wave of AI adoption will not stay inside engineering. Brian Kyed, one of the founders, put it bluntly: there will always be a newer model, so the safer move is to make the model swappable and the company context durable. That is a strong argument in a market where model churn is already a feature, not a bug.

Palette Desktop is the part customers can use today. It lets teams work with agents such as Claude Code and Codex directly on their files and folders, with a choice between cloud models for speed, capability and cost, or local models when more control over sensitive information is needed. The broader Palette OS is still with design partners, and it adds a company-owned context library, shared AI skills and a connector gateway meant to plug tools in once and reuse them across multiple agents.

The fresh money will go into sales and hiring, alongside continued product development. That is the practical part of the story, and probably the harder one: getting companies to trust an AI layer that is supposed to sit above the models, not become another brittle tool employees forget to use after the demo.