How a Gates Foundation-Backed A.I. Startup Tackles Weather Forecast in Africa: CEO Q&A

Africa’s agricultural sector is grappling with a major challenge: a severe lack of reliable atmospheric data, worsened by the impacts of climate change. This shortage of data has led to inaccurate weather forecast, poor crop yield and wasted resources, especially in regions like sub-Saharan Africa. Agriculture is the lifeblood of many African nations, supporting more than half of the continent’s labor force. Yet, productivity has fallen 34 percent since 1961 due to climate change, more than any other region in the world, according to a 2023 report by the World Meteorological Organization (WMO). This decline creates a ripple effect, threatening food security, destabilizing ecosystems and damaging economies. Accurate weather prediction can be a game-changer for the continent and its farmers, helping them make smarter decisions, reduce financial losses and improve food security. Africa’s economic losses due to climate change are expected to range between $290 billion and $440 billion, depending on how much temperatures rise, according to the African Economic Outlook, an annual report published by the African Development Bank (AfDB). Despite the urgency, funding for climate adaptation remains woefully inadequate. But there’s hope on the horizon. WindBorne Systems, a Palo Alto, Calif.-based A.I. startup, is determined to transform how weather data is captured and utilized. The company was founded in 2019 by John Dean, a former engineer at SpaceX and Lyft (LYFT) building autonomous systems, through a startup program under Stanford University, where Dean dropped out as a graduate student to focus on the company. WindBorne aims to tackle climate change by improving the accuracy of weather prediction using A.I.-powered weather balloons that gather atmospheric data. The startup claims its A.I. model outperforms competitors like Huawei’s Pangu and Google DeepMind’s GraphCast, as well as major physics-based models like GFS and IFS. 

So far, WindBorn has raised a total of $25 million from Khosla Ventures, Footwork VC and Susa Ventures. Its clients include the National Oceanic and Atmospheric Administration (NOAA) and the U.S. Air Force. WindBorne also has a partnership with the Bill and Melinda Gates Foundation, which has invested $5 million in the startup to develop A.I. weather forecasting technologies and A.I. models tailored for smallholder farmers in sub-Saharan Africa.

In November, Observer interviewed Dean about how WindBorne tackles one of Africa’s most pressing climate challenges while boosting agricultural productivity, and how A.I.-driven solutions could play a pivotal role in the global fight against climate change.The following conversation has been edited for length and clarity.

Observer: Why does WindBorne focus its initial efforts on Kenya, and how does this region exemplify the broader challenges facing Africa?

John Dean: Kenya was a natural choice for our first area of operation due to its combination of significant challenges with data scarcity and a relatively advanced meteorological infrastructure. Our data is being used to support efforts in renewable energy integration, disaster preparedness and carbon capture—all contributing to reversing climate change globally. Through partnerships with the Kenya Meteorological Department (KMD), WindBorne addresses the data gap that has long hindered accurate weather forecast in the region.Precise weather prediction significantly impacts smallholder farmers, who rely heavily on rainfall. For example, accurate forecast can inform farmers about the optimal time to plant or irrigate crops, maximizing yields and conserving resources. This can prevent costly mistakes like planting too early or wasting water before an imminent storm. How does WindBorne’s Global Sounding Balloons (GSBs) address Africa’s weather data scarcity? How are they different than conventional weather balloons?

They gather critical weather data across the troposphere. Each WindBorne balloon weighs only 2.5 kilogram, equipped with proprietary sensors and advanced avionics, allowing for real-time data collection and precise navigation. Unlike other weather systems, GSBs can collect 150 times more data per dollar over land and 250 times more over oceans. Moreover, the new version of GSBs can fly autonomously for over 50 days—significantly longer than the few hours traditional radiosondes can operate. 

The collected data is fed into A.I. models, enabling hyper-local, actionable weather insights. Our innovation can potentially transform sectors such as agriculture, transportation and energy management across the continent and on a global level, providing more reliable forecasts for crop planning to energy resource management.

What makes WindBorne’s WeatherMesh A.I. model stand out in the competitive field of weather forecasting?

WeatherMesh AI achieves high accuracy through three main factors: vision transformer-based deep learning, advanced in-house modeling techniques, and the integration of live atmospheric data from our observations. We custom-built a computing infrastructure that enables us to run these complex models effectively. Since its launch in February, we’ve been integrating unique atmospheric data into the model, leading to ongoing improvements in accuracy. A.I.-driven models are faster and far more computationally efficient than conventional weather models. WeatherMesh AI combined with GBS’s ability to adjust altitude for targeted measurements ensures precise insights while being cost-effective. 

How is the Gates Foundation partnership advancing WindBorne’s mission in Africa?

The partnership enables collaboration with local meteorological agencies like KMD, ensuring that captured data is integrated into regional forecasts and meets the needs of local communities. Moreover, our local teams are also working closely with the KMD team to support their technical capabilities. So far, over 70 balloons have been launched from KMD’s Nairobi headquarters. 

Next year, we plan to establish a second launch site in the region to collect even more observations. The partnership helps us gather direct feedback and refine our products to meet each region’s unique needs. Being on the ground and collecting live input is essential. As an end-to-end weather solution, we handle everything from gathering in-situ data to integrating it into models and collaborating with various stakeholders to build a comprehensive system that improves forecast accuracy and supports smallholder farmers across the region.

By proving the success of our operations, we aim to scale our technology across other African nations and broader developing countries. 

What broader impacts could WindBorne’s technology have on Africa’s rural economies and beyond?

Our expansion into Sub-Saharan Africa aims to address critical challenges in crop yield optimization, food scarcity and GDP gaps, particularly in regions like Kenya, where losses on farms are nearly double the global average.

We are collecting key precipitation data to improve future forecasts by launching balloons during Kenya’s rainy season. Over 2025, we will continue collecting in-situ observations, working with local agencies like KMD and IGAD to integrate the data into their models and WindBorne’s A.I. systems. These improved forecasts will help farmers optimize planting schedules and water usage, ultimately enhancing crop yields.Beyond agriculture, industries such as renewable energy, disaster response, aviation and construction stand to benefit from more accurate weather data. For instance, energy grids can optimize renewable resource integration while shipping routes can be planned to save fuel and reduce emissions. 

What challenges are you facing to scale current operations in Africa?

Expanding to remote African areas involves logistical hurdles, such as infrastructure gaps and navigating regulatory requirements for airspace use, as approvals vary widely across countries. We aim to address these challenges by collaborating closely with local stakeholders and ensuring its technology remains cost-effective and sustainable.

To ensure sustainability and affordability, we will continue to strengthen our collaborative partnerships and work with global organizations like the Gates Foundation and local entities like KMD, we can leverage existing resources and it helps disseminate our technology more effectively and affordably.

At the core, WindBorne’s designs are modular and scalable, enabling us to set up a site and collect observations at a fraction of the cost of existing solutions. Furthermore, for each site, we hire a local balloon launcher. Equipping local teams to handle operations ensures the technology deployment is cost-effective and sustainable while fostering ownership within the communities we operate.

How does WindBorne’s open-source data approach empower local communities, and could it redefine how tech companies foster innovation and collaboration in underserved regions?

By making data easily accessible, we’re fostering collaboration and innovation, working with local researchers, academics and private companies to create solutions tailored to Africa’s challenges. This approach ensures that all voices are heard in decision-making, leading to more inclusive and impactful solutions.