U.S. Air Force Taps WindBorne Systems for Global Atmospheric Data in Multi-Million Dollar Deal

The U.S. Air Force has entered into a significant new partnership with Redwood City-based startup WindBorne Systems, securing a contract that could be worth up to $49 million. This agreement marks a major milestone for the atmospheric technology company, as it seeks to scale its fleet of autonomous weather balloons and further integrate its proprietary, AI-driven forecasting capabilities into national security operations.

Under the terms of the deal, the Air Force will procure high-resolution atmospheric sensor readings collected by WindBorne’s expansive constellation of weather balloons. Currently, the company maintains approximately 700 of these autonomous vehicles hovering globally. The data harvested from these flights serves a dual purpose: it is sold to a variety of commercial and government clients and serves as the primary empirical fuel for WindBorne’s internal deep-learning weather models.

While WindBorne has previously cultivated relationships with federal entities, including the National Weather Service, the U.S. Navy, and the Air Force itself, this contract represents the most substantial financial commitment the startup has received to date. The structure of the agreement is designed for operational flexibility; the $49 million figure represents a ceiling rather than an upfront payment. Instead, the Air Force operates on a "pay-as-you-need" model, allowing military planners to place orders for data collection in specific geographic regions during critical timeframes. To date, the Air Force has already placed $1.7 million in orders, with company leadership anticipating that demand will continue to grow as the efficacy of the platform is proven in field conditions.

Scaling Through Strategic Partnerships

For WindBorne, the contract is more than a simple service agreement; it is a catalyst for hardware and software advancement. According to Andrey Sushko, the company’s co-founder and Chief Technology Officer, the revenue generated from this deal provides a stable foundation to expand their constellation and refine the underlying technology.

"This is largely a vehicle through which continued growth of the constellation can be supported, as well as some work on research and development to continue to improve the hardware and performance of the balloons," Sushko explained.

The agreement is notable for being awarded without competitive bidding, a testament to the niche, high-performance capability WindBorne has carved out in the weather-tech sector. Sushko noted that the government’s decision reflects a determination that, for the specific requirements of their global operations, WindBorne is currently the only provider capable of delivering the necessary atmospheric data at the required scale and precision.

This government validation comes on the heels of a successful $37 million Series B funding round reported by CEO John Dean in August. The influx of capital is intended to support the company’s aggressive scaling strategy. Earlier this summer, the firm unveiled "WeatherMesh 6," the latest iteration of its forecasting model. The company claims that WeatherMesh 6 consistently outperforms traditional numerical weather prediction models, such as those maintained by the European Centre for Medium-Range Weather Forecasts (ECMWF). While traditional models rely on simulating atmospheric physics through massive, static equations, WindBorne’s model utilizes deep learning to interpret data in real-time, offering a new paradigm for meteorological accuracy.

The Rise of AI in Meteorology

The field of weather forecasting is currently undergoing a radical shift, driven by the emergence of transformer-based architectures and the availability of massive historical datasets. For decades, meteorology was the exclusive domain of supercomputers running complex, expensive physical simulations. These systems, while powerful, were often limited by the sheer computational power required to process granular atmospheric changes.

The integration of deep learning has fundamentally changed this equation. By ingesting decades of historical weather data, AI models can now identify patterns and make predictions with significantly greater speed and, in some cases, improved accuracy compared to traditional methods. While traditional physics-based models remain a cornerstone of global meteorology, AI is increasingly viewed as an essential complement that fills the gaps in our understanding of complex, fast-moving weather events.

The effectiveness of this hybrid approach is becoming evident across the industry. Frontier AI labs and government research agencies are reporting that their models are beginning to eclipse the performance of legacy systems. Furthermore, private-sector innovators like Google DeepMind have introduced advanced weather models that are rapidly becoming central to disaster response strategies. In the United States, these tools are playing an increasingly critical role in how the government prepares for and responds to the hurricane season, providing emergency managers with more reliable data to protect lives and infrastructure.

Operational Success at the Extreme

WindBorne’s ability to command such high-level interest is rooted in its unique operational achievements. In August, the company demonstrated the capabilities of its technology in a high-stakes environment by successfully maneuvering an unpowered weather balloon into the eye of Typhoon Dolphin.

This feat of aerial navigation is achieved through sophisticated flight control. WindBorne’s balloons do not have engines; instead, they maintain a specific course by adjusting their altitude. By bleeding air to descend or dropping sand ballast to ascend, the balloons can catch different wind currents at various strata of the atmosphere, effectively "steering" the vehicle to a target destination.

"That’s a pretty novel step, and the two ingredients for that are one, having good wind and weather models, and knowing how to navigate them; and two, having a large enough constellation that you have balloons in the right places," Sushko said, reflecting on the milestone. "It’s one of the first times that a deep-learning model was used for actual operational decisions."

By placing these sensors directly into the heart of extreme weather events, WindBorne is providing data that was previously impossible or prohibitively dangerous to collect. This granular, real-time data is precisely what makes the company’s AI models so potent, as they are trained on the very empirical measurements they are designed to predict.

As the company looks to the future, the partnership with the U.S. Air Force serves as both a validation of its technology and a critical revenue stream to fund further innovation. Whether through improving the endurance of its balloon fleet or refining the predictive power of its WeatherMesh models, WindBorne is positioning itself at the center of a technological revolution in meteorology. As the frequency and severity of extreme weather events continue to impact global stability and national security, the ability to predict the atmosphere with high precision—and at a lower cost than traditional supercomputing—will only become more essential. For now, the "pay-as-you-need" arrangement ensures that the Air Force has a front-row seat to these advancements, providing a template for how private-sector AI innovation can be integrated into the machinery of national defense.