WindBorne Systems has raised $37 million in Series B funding as the weather technology startup looks to transform more accurate artificial intelligence forecasts into products that businesses and government agencies will pay to use.
The financing was co-led by Khosla Ventures and Galvanize, with participation from TransLink Capital, Lux Capital, previous investors, and other backers. The deal gives the company a post-funding valuation of approximately $250 million.
Founded in 2019, WindBorne combines artificial intelligence forecasting models with atmospheric measurements collected by a global network of long-duration weather balloons. The company says the combination gives it access to valuable information from oceans, remote regions, and other locations where conventional weather observations remain limited.
WindBorne currently operates from 20 launch sites around the world and has approximately 600 balloons in flight at any given time.
Unlike traditional weather balloons, which generally remain active for around two hours and produce a single vertical atmospheric profile, WindBorne’s autonomous balloons can remain airborne for several weeks. According to the company, each balloon can generate around 50 controlled vertical profiles during its mission.
These profiles measure conditions such as air temperature, pressure, humidity, wind speed, and wind direction at different altitudes. The information can then be added to forecasting models to produce a more detailed picture of changing atmospheric conditions.
The scale of the data gap remains significant. Traditional radiosondes are released at 92 locations across the United States at least twice each day. Globally, around 900 radiosondes are launched at each of the two standard observation times, resulting in roughly 1,800 launches per day.
However, many of those launches occur from land. Large sections of the world’s oceans and remote territories have fewer direct observations. WindBorne is attempting to fill these gaps with balloons that can travel long distances while repeatedly moving through different atmospheric layers.
The company is also developing sensor packages that can descend into the ocean and continue operating as floating measurement devices. This could allow one mission to collect information from both the atmosphere and the sea.
Weather forecasting has traditionally relied on physics-based numerical models that require large amounts of computing power. Running detailed global simulations has usually been limited to national weather agencies, research organizations, and companies with access to major supercomputing infrastructure.
Recent advances in deep learning have started to change that calculation. AI forecasting systems can process large historical and real-time datasets while operating with lower computing requirements than many traditional simulation methods.
WindBorne uses its proprietary balloon observations alongside data produced by government weather agencies. Its WeatherMesh forecasting system is designed to combine these information sources and generate predictions more quickly.
The company believes its privately collected observations provide an advantage that competing AI models cannot easily reproduce. While forecasting algorithms may become widely available, building and operating a global network of hundreds of long-duration balloons requires specialized hardware, launch infrastructure, flight-control software, and regulatory experience.
The new funding will be used partly to expand computing capacity and improve communication between balloons. WindBorne plans to reduce its reliance on satellite communications by developing a mesh radio network that would allow balloons to exchange information with one another.
Government organizations currently represent a large portion of WindBorne’s customer base.
The U.S. National Weather Service purchases atmospheric data from the company. WindBorne also has research partnerships with the U.S. Air Force and U.S. Navy.
One area of development involves forecasting systems that can operate directly aboard naval vessels. A model installed on a ship could continue producing local predictions even when its connection to land-based computing systems is limited or temporarily unavailable.
Government demand gives WindBorne an established source of revenue while the company develops its private-sector business. Public agencies already have teams, software, and operational processes built around weather information, making it easier for them to use new data sources.
Commercial customers may require more than access to a forecast. They often need the information converted into a specific recommendation, financial signal, operational alert, or automated decision.
WindBorne is now targeting industries where weather can have a direct financial impact.
Investment funds can use forecasts to evaluate agricultural production, energy demand, transportation disruptions, and commodity prices. Airlines may use more detailed predictions to support routing and de-icing decisions, while shipping companies can adjust routes to reduce fuel consumption or avoid severe conditions.
Energy companies can use temperature and wind predictions to estimate electricity demand and renewable power generation. Retailers, insurers, construction businesses, logistics operators, and agricultural companies could also benefit from forecasts connected directly to operational planning.
The challenge is turning raw atmospheric data into a product that fits existing business systems.
Several companies have built sophisticated Earth-observation networks but struggled to develop large commercial markets. Government customers are accustomed to purchasing scientific data, while many businesses want a completed decision-making tool rather than another dataset that requires specialist interpretation.
WindBorne plans to use part of the $37 million funding round to expand its sales and commercial product teams. The goal is to make forecasts easier to integrate into business software and automated workflows.
WindBorne’s opportunity depends on whether improved accuracy can produce measurable financial value.
A slightly better forecast may be important to a meteorologist, but a commercial customer needs to understand how that improvement affects revenue, costs, safety, or operational risk. A shipping company might measure fuel savings, while an energy trader may focus on changes in demand forecasts. An airline could calculate reductions in delays or unnecessary de-icing procedures.
AI may help bridge the gap by connecting weather predictions to these specific outcomes. Instead of presenting customers with maps and raw measurements, future systems could recommend when to reroute a shipment, adjust an energy position, protect crops, or change staffing levels.
With a $250 million valuation, 20 global launch sites, and around 600 balloons continuously collecting atmospheric information, WindBorne has assembled a substantial observation network.
Its next challenge is commercial rather than scientific. The company must prove that more accurate weather intelligence can become a repeatable and profitable business tool across industries.
Share your thoughts about this article.
Be the first to post a comment!
