WindBorne’s Planetary Nervous System Takes Shape
Two years ago we wrote that WindBorne sat at the forefront of a once in a generation revolution in weather forecasting. This month, we’re excited to report that the market agrees.
We’re pleased to share that WindBorne Systems has closed an oversubscribed $37 million Series B, co-led by Khosla Ventures and Galvanize, with participation from Translink Capital, Biprogy, Lux Capital, and existing investors. The round brings WindBorne’s total funding to more than $62 million, and it follows a year in which the Company tripled revenue and tripled the size of its balloon constellation.
When we joined WindBorne’s Seed round, the thesis was simple: traditional weather data collection was falling behind a changing climate and a vertically integrated platform that owned its own observations could out forecast anyone. At the time, the balloons flew for up to a month and the forecast accuracy gains were measured against models that had barely changed in decades. The bet was on engineering depth and a team we had known for over a decade.
What has changed since then is scale, and with it, proof.
From Better Data to Best Weather Model
The most important development of the past year isn’t the round itself. Earlier this year, WindBorne released WeatherMesh 6, which now sits at the top of publicly available forecasting benchmarks, ahead of both the leading AI weather models and the traditional physics based systems run by government agencies. That is the vertical integration thesis playing out exactly as designed: the balloons gather observations from parts of the atmosphere nobody else is sampling, and those observations feed a proprietary model that compounds in accuracy as the network grows.
WindBorne’s data is now assimilated into NOAA’s Global Forecast System, the public backbone behind much of the consumer weather infrastructure in North America. A growing share of the forecasts people see every day already carry WindBorne’s fingerprints.
Who is Buying
When we first invested, we pointed to investment funds and insurance carriers as the customers who would pay first for a longer, more accurate forecast horizon. Trading firms were indeed among the earliest commercial adopters and demand from government agencies has grown remarkably over the past two years — by a factor over close to 10X. The Series B capital is earmarked for expanding the sensing network and the compute behind WeatherMesh. That’s just more fuel for the flywheel here.
Why This Still Matters
Our original excitement was about cost: a platform that could eventually drive the price of atmospheric data low enough that nearly any customer could build proprietary weather applications on top of it. A constellation of close to 1000 balloons, a benchmark leading model, and a revenue base that tripled in twelve months is what that trajectory looks like from the inside. Extreme weather is only getting more expensive, and the value of knowing what’s coming a few days earlier is being repriced upward across energy, agriculture, infrastructure, and emergency response.
Congratulations to John Dean, Dr. Andrey Sushko, Kai Marshland, and the entire WindBorne team. We said this was one of the most technically complex but rewarding opportunities we had seen. It still is, and it’s getting even bigger.
Read WindBorne’s full announcement here.