Atmos's $200 Million Design Pipeline Couldn't Outrun the Fed

The YC and Khosla-backed custom home marketplace shut down last year after 50 homes and $7 million in revenue, a casualty of interest rate hikes.

About Atmos

Published

The math was simple. A custom home designed through Atmos would cost roughly $225 per square foot, a price point aimed at the Sun Belt's growing families [Boring Business Nerd, ~2023]. The promise was simpler still: an end-to-end digital platform to manage the entire process, from lot selection to builder matching. By late 2022, the startup had designed $200 million worth of projects. Then the Federal Reserve raised interest rates, and the pipeline froze solid [TechCrunch, Dec 2025].

Atmos, a San Francisco-based proptech marketplace founded in 2019, ceased operations in March 2025. It had raised nearly $20 million from investors including Y Combinator, Khosla Ventures, and Sam Altman [Startups.RIP, ~2025]. It completed about 50 homes and generated approximately $7 million in revenue [TechCrunch, Dec 2025].

The wedge: from VR to transactions

Atmos began with a visualization tool. Co-founders Nicholas Donahue, Trent Hedge, Matthew Rastovac, and Austin Kahn, all recognized on the Forbes 30 Under 30 list in 2020, initially built virtual reality software to help clients envision their homes [Forbes, Dec 2022] [Raleighnc.gov, ~2020]. The team, which included Donahue, who grew up in a homebuilding family, quickly pivoted [Topio Networks, retrieved 2026]. The real wedge, they determined, was not in showing the dream but in managing the notoriously fragmented transaction to build it.

The company evolved into a managed marketplace. It served the Charlotte and Raleigh metro areas, offering personalized guidance on land due diligence, floor plan design, and builder selection [buildatmos.com, retrieved 2026]. For builders, it promised a stream of vetted, ready-to-build leads. For buyers, it promised a single point of contact and a managed workflow to replace a chaotic process involving architects, contractors, and lenders.

Capital and credibility

Investors bought the thesis that software could bring order to custom homebuilding. A $4 million seed round in 2020 was followed by a $12.5 million Series A in November 2022 led by Khosla Ventures [TechCrunch, Nov 2022]. The cap table also included Bedrock, JLL Spark, and Figma CEO Dylan Field.

The funding supported a team that peaked at 40 employees and the development of a platform intended to handle everything from site planning to financing [TechCrunch, Dec 2025].

Round Amount
Seed (2020) $4M
Series A (2022) $12.5M

Where the model met the market

The company's traction highlights both its early promise and its ultimate constraints. Designing $200 million in projects is a significant indicator of demand and sales activity. Translating that into only 50 completed homes and $7 million in revenue, however, reveals the fatal gap in the model: the lengthy, capital-intensive construction cycle.

  • Project velocity. Completing roughly 50 homes over several years points to a high-touch, slow-turn operation.
  • Revenue concentration. With an average revenue per home of approximately $140,000 (estimated), the business was reliant on a small number of very large, complex transactions.
  • Macro sensitivity. The entire pipeline was hostage to mortgage rates and buyer confidence, dependencies that pure software companies often insulate against.

The interest rate reckoning

The most credible risk to Atmos's model was always macroeconomic, and it materialized decisively. The Federal Reserve's series of interest rate hikes beginning in 2022 directly attacked the affordability of the custom homes Atmos was selling. Projects in the design pipeline were paused or canceled. The $200 million in designed value became largely theoretical almost overnight [TechCrunch, Dec 2025].

This was not a failure of product-market fit in a stable environment. Demand was clearly present. It was a failure of business-model fit for a volatile economic climate. A capital-efficient marketplace might have weathered the storm, but a capital-intensive, project-managed service with fixed operational costs could not. By March 2025, the board made the decision to shut down.

The founder's second act

CEO Nicholas Donahue did not wait long to pivot. Within months of Atmos winding down, he launched Drafted, an AI-powered floor plan design startup. He took four former Atmos employees with him and quickly raised $1.65 million at a $35 million post-money valuation [TechCrunch, Dec 2025]. The new venture applies lessons from the first. It is purely a software product, avoiding the managed services and physical asset risks that doomed Atmos. It is a bet on automating the front-end design, not owning the entire back-end build.

Read on Startuply.vc