2 min read AI-generated

$34 million for Paper: the design platform for the agentic era

Copy article as Markdown

Paper wants to close the gap between design and code — for teams building software with AI agents. Accel and ICONIQ lead the round, and the angel list reads like a who's who of Anthropic and OpenAI.

Featured image for "$34 million for Paper: the design platform for the agentic era"

While everyone watches the big model launches, a second layer is forming underneath: the tools that turn AI agents into actual productivity. A fresh example — Paper closed a $34 million Series A on July 23, led by Accel and ICONIQ.

What it’s about

Paper calls itself a «design platform for the agentic era». The idea: design and production code should no longer be two separate worlds you constantly translate between. Paper renders directly with HTML and CSS, so designs connect straight to the real code — and the AI agents writing that code sit at the same table. Instead of «designer throws a mockup over the fence, engineer rebuilds it», humans and agents work on the same surface.

Paper was founded by Stephen Haney, a game developer who previously ran engineering at the developer platform WorkOS. The company says annualized recurring revenue has grown 25x since launch, landing it on Ramp’s fastest-growing-companies list four months running. Users include Ramp, Lovable, Vercel, PostHog and Y Combinator.

The interesting angel list

What makes this round special for us here: the angels. Alongside Michael Grinich (WorkOS) and Anton Osika (Lovable), engineers and designers from Anthropic and OpenAI are personally invested. That’s no accident — it shows where the people building these models think the next value lies: not just in the model itself, but in the layers above it that make the model usable.

My take

To me this is the genuinely interesting story behind the headlines. A model like Opus 5 is the engine — but an engine alone doesn’t drive anywhere. It needs the chassis, the steering wheel, the workshop around it. Paper is exactly that kind of chassis: the connection between what you imagine and what the agent builds from it. As someone who spent years bouncing between design and frontend, I know how much friction lives at that seam. If AI agents take that friction away, it’s not a niche topic — it’s the everyday work of tomorrow.


Sources: