Sky Sensors: WindBorne Secures $37M to Profit From Weather Prediction Models

Deep learning techniques are reshaping how experts predict global weather patterns. Modern forecasting models run directly on basic laptops rather than massive supercomputers, lowering technical hurdles across the meteorology industry. Weather startup WindBorne Systems collected $37 million in Series B funding to turn those cheap software forecasts into a profitable commercial enterprise. CEO John Dean confirmed that the fresh funding round values the firm at $250 million post-money.

Khosla Ventures and Galvanize co-led the financing round. TransLink Capital, Lux Capital, and existing investors also participated in the transaction. Founded in 2019, WindBorne built cheap weather sensors attached to long-duration weather balloons. The rapid growth of machine learning prediction tools over the past four years allowed private companies to process atmospheric data directly, bypassing expensive government supercomputers.

WindBorne currently operates 20 launch sites globally, keeping roughly 500 balloons floating in the atmosphere at any given moment. These balloon-mounted sensors collect critical atmospheric data from remote, hard-to-reach locations, including active storm centers like the eye of a typhoon. Moving forward, the team plans to deploy ocean-bound sensor packages that drop into open water to gather continuous sea surface metrics.

The firm feeds these proprietary balloon metrics into its primary forecasting system, combining private observations with public datasets from government weather agencies. Dean calls this global data collection effort a planetary nervous system. He pointed out that adding physical balloon sensors to standard forecasts increases accuracy far better than relying solely on satellite images. Proving that private buyers will pay for higher accuracy helped de-risk the business for venture capital backers.

Government agencies make up WindBorne primary customer base today. The National Weather Service buys environmental data directly from the startup, while the Air Force and Navy fund research partnerships to build localized forecasting tools for military vessels operating with spotty internet connectivity.

Beyond public contracts, WindBorne wants to expand commercial sales across the private sector. Financial investment funds buy custom weather data to predict commodity price shifts, agricultural yields, and energy demands. The startup is using its fresh capital to build out dedicated sales teams, replace satellite links with mesh radio networks, and sign private corporate accounts.

Selling raw weather data to private buyers remains difficult. Over the last decade, several earth-observation startups struggled to scale because commercial clients found it hard to integrate raw environmental metrics into daily business operations. Private weather firms usually repackage public government forecasts for news stations, flight routing, or commercial shipping lines.

Saloni Multani, a partner at Galvanize who co-led the funding round, explained that integrating weather metrics into broader business decisions was historically too expensive for private firms. However, better AI models make processing environmental data simple and cost-effective. By linking accurate predictions to real business outcomes, WindBorne aims to build a scalable, highly profitable commercial weather enterprise.