Forward Deployed Operator

Why the US Air Force wants data from A startup's 700 floating balloons. By Abdul Wasay | 4 hours ago | WindBorne Systems has won a contract worth up to $49 million from the US Air Force. The government awarded this deal without competitive bidding because it determined that WindBorne was the only company that could meet its data collection needs. As a result, this contract also marks the largest in the company's history. However, the Air Force has only ordered $1.7 million in services so far under its pay-as-needed structure. The deal specifically involves the 557th Weather Wing and the 16th Weather Squadron. Under this arrangement, WindBorne will supply weather data from its fleet of roughly 700 autonomous balloons. These patented balloons weigh under 2 kilograms each and can stay airborne for over 50 days, with the company's record standing at 104 days. In addition, one balloon has completed a double circumnavigation of the globe in a single flight. WindBorne was founded at Stanford University and is now based in Palo Alto, California. CEO John Dean and CTO Andrey Sushko lead the company. Furthermore, the startup raised $37 million in its Series B round at a $250 million valuation. Khosla Ventures and Galvanize co-led that round, while TransLink Capital and Lux Capital also participated. The balloons steer themselves by releasing sand ballast or venting gas to ride different wind currents. Because of this design, they collect up to 150 times more data per cost dollar than conventional weather balloons. Moreover, each flight records dozens of vertical atmospheric profiles measuring temperature, wind speed, pressure, and humidity. The World Meteorological Organization notes that 85% of the atmosphere remains inadequately measured, so this data fills a critical gap. On top of that, WindBorne released its WeatherMesh 6 AI model in early summer 2026. This transformer-based system generates a 10-day forecast in under 60 seconds on a single GPU. That makes it over 1,000 times faster than traditional physics-based models. It also achieves 3-kilometer resolution across the US and Europe. Chief Product Officer Kai Marshland says the model is "as accurate five days out as a traditional forecast is the day before." Additionally, it performs up to 30% better than the European Centre for Medium-Range Weather Forecasts model and 11% better than Google DeepMind's GraphCast. In August 2026, WindBorne became the first company to send an unpowered flight into a typhoon's eye. Sushko called this mission "one of the first times that a deep-learning model was used for actual operational decisions." Similarly, the company had previously sent six balloons through Hurricane Milton in October 2024. Meanwhile, WindBorne has also held a separate Navy contract since July 2024 to extend WeatherMesh into subseasonal forecasting. Beyond the military, WindBorne has served NOAA as an operational customer since early 2023. The company also works with the National Center for Atmospheric Research and the Defense Innovation Unit. Furthermore, it now sells forecasts to investment funds that predict commodity prices. Governments worldwide spend roughly $10 billion per year on weather observation, and WindBorne estimates the total addressable forecasting market at $100 billion. The company also plans to scale to 10,000 balloons by 2028 with around 300 launches per day from approximately 30 sites. It is currently replacing satellite communications on its balloons with a mesh radio network. On top of that, WindBorne has started dropping aerial sensor packages that become floating ocean buoys after descent. Dean calls the resulting dataset a "planetary nervous system."

WindBorne wins a hefty US Air Force contract for its balloon data. Published: 2026-10-12 18:19 | Source: TechCrunch " AI A Silicon Valley startup that uses weather balloons and AI to predict the weather won a U.S. military contract worth almost $50 million. The U.S. Air Force has awarded WindBorne Systems a contract worth up to $49 million to buy data collected by the company's fleet of autonomous weather balloons. WindBorne Systems currently has around 700 weather balloons flying globally, collecting atmospheric sensor readings that are sold to customers and used to feed the company's AI weather forecasting model. The contract is the largest the company has participated in and will support the continued growth of the constellation and research and development to improve the hardware and performance of the balloons. WindBorne Systems has developed a deep-learning weather forecasting model called WeatherMesh 6, which provides more accurate forecasts than traditional models. The company has worked closely with the U.S. government and has become the first organization to send an unpowered flight system to collect data in the eye of a typhoon. The use of AI in weather forecasting has become more prevalent in recent years, with models built by frontier labs and government agencies proving more accurate than traditional forecasts. WindBorne Systems' contract with the U.S. Air Force is a significant development in the field of AI weather forecasting, as it demonstrates the potential for AI to be used in operational decisions. The company's ability to maneuver its balloons using wind and weather models has also enabled it to collect data in previously inaccessible areas, such as the eye of a typhoon. The contract is expected to support the continued growth of WindBorne Systems and the development of its AI weather forecasting capabilities. The company has raised $37 million in a Series B round to continue scaling its business and has released the latest version of its forecasting model, which provides more accurate forecasts than traditional models. The use of AI in weather forecasting has the potential to revolutionize the field, enabling more accurate and efficient forecasting. WindBorne Systems' contract with the U.S. Air Force is a significant step towards this goal, demonstrating the potential for AI to be used in operational decisions and enabling the collection of data in previously inaccessible areas. The company's deep-learning weather forecasting model, WeatherMesh 6, is a key component of this effort, providing more accurate forecasts than traditional models. The contract is also expected to support the continued growth of WindBorne Systems, enabling the company to further develop its AI weather forecasting capabilities and expand its operations. In the future, the use of AI in weather forecasting is expected to become even more prevalent, with models built by frontier labs and government agencies continuing to improve in accuracy and efficiency. WindBorne Systems is well-positioned to play a key role in this development, with its contract with the U.S. Air Force demonstrating the potential for AI to be used in operational decisions. The company's ability to collect data in previously inaccessible areas, such as the eye of a typhoon, is also expected to enable new insights and discoveries in the field of weather forecasting. Overall, the contract between WindBorne Systems and the U.S. Air Force is a significant development in the field of AI weather forecasting, demonstrating the potential for AI to be used in operational decisions and enabling the collection of data in previously inaccessible areas. The company's deep-learning weather forecasting model, WeatherMesh 6, is a key component of this effort, providing more accurate forecasts than traditional models. The contract is also expected to support the continued growth of WindBorne Systems, enabling the company to further develop its AI weather forecasting capabilities and expand its operations. No more hand-written code" or AI-free art for delayed Marvel/Star Wars shooter. Russian bloggers warn Russia's hyper-centralized digital economy is vulnerable. Learn how to run OpenAI GPT-6 Sol, Luna, and Astra on Amazon Bedrock through global cross-Region inference from Singapore, Jakarta, Thailand, Malaysia, and Taipei. This post covers the Converse and OpenAI-compatible APIs, IAM and SCP permissions, quota sizing, and monitoring inference from your source Region. Learn how to deploy decision models with llama.cpp on Amazon SageMaker AI. This technical how-to serves OpenAI-compatible chat completions with the AWS Deep Learning Container for llama.cpp, then extends the runtime with SemIf to return probabilities over declared options on CPU or GPU. Published: 2026-10-12 16:45 | Source: TechCrunch " AI Journalists, entrepreneurs, and even some public relations professionals will tell you that the PR process is broken. But could AI actually make things better? Published: 2026-10-12 16:27 | Source: TechCrunch " AI Blueprints AI, a bootstrapped startup based in Santa Monica, developed software that produces code-compliant construction blueprints from a photo, PDF, or sketch.