Drafted lands $16 million seed led by Buckley Ventures
Category: Startups
By James Whitemore
Published: 2026-06-01T09:07:24.000Z
AI keeps creeping into industries that felt safe from it, and home design is next. Drafted, a startup rebuilding custom design around generative models, has raised a $16 million seed led by Buckley Ventures. The pitch is blunt. Most people live in homes they never got to shape.
AI keeps creeping into industries that used to feel safe from it, and residential design is the latest one in the crosshairs. Drafted, a startup trying to rebuild custom home design around generative models, has raised a $16 million seed round led by Buckley Ventures, the company said in late may. The pitch is blunt. Most people live in homes they had almost no say in shaping, and Drafted thinks software can finally change that. The company is led by Nick Donahue, who has been chasing this exact problem for years. His previous venture, Atmos, leaned on human designers backed by software, raised around $20 million from investors including Khosla Ventures and Sam Altman, and designed roughly $200 million worth of homes before shutting down in early 2025. The killer was timing, since rising interest rates left buyers unable to finance the houses they had spent months designing. Donahue describes that effort as a glamorized architecture firm, and Drafted is his answer to its weaknesses. This time there are no in house designers, just an AI model that returns floor plans and exterior concepts in minutes. The economics are the interesting part. Hiring an architect for a custom home can run anywhere from $10,000 to well past $50,000 and take months, while ready made template plans are cheap but rigid. Drafted aims for the middle, with basic plan packages starting near $1,000. Its proprietary model was trained on real house plans that actually passed permitting and got built, which the company says keeps the output grounded in practical constraints. Generating a single design costs around two tenths of a cent, far less than leaning on a general purpose model. Users punch in details like lot size, room count and square footage, describe the style they want, and get options they can keep refining. The early traction helps explain the raise. Over a single recent month, roughly 120,000 people generated more than 325,000 designs, mostly through social sharing. The roughly nine person team includes researchers from Stanford, Autodesk, Brown and Adobe, plus architects from WeWork and Atmos, and the backer list runs from Y Combinator and Patrick Collison to Pinterest co-founder Ben Silbermann and musician Ryan Tedder. Total funding since launch sits around $17.5 million. The Middle East is one of the more natural markets for a tool like this. Proptech is booming across the Gulf, with the UAE sector projected to nearly triple to about 5.69 billion dirhams by 2030 and Saudi Arabia leaning hard into property tech through Vision 2030 and giga projects like NEOM and Roshn that target hundreds of thousands of new homes. The region also moves fast, since policymakers, developers and builders are often aligned, letting new tools jump from pilot to scale quicker than in slower legacy markets. For an AI design platform, that combination of demand and speed is exactly the opening.