Jua

Jua is a Swiss startup based in Zurich that has developed an artificial intelligence (AI) platform aimed at revolutionizing weather forecasting and climate modeling. The company has trained what it claims to be the world's first global high-resolution weather prediction model using an end-to-end deep learning approach and petabytes of data from tens of millions of sensors.

Jua's AI model promises highly accurate weather and climate forecasts with capabilities such as 16-day forecasts and precision surpassing traditional numerical models. It can provide global coverage with a resolution of one square kilometer every five minutes, while some regions are covered at a resolution of up to 250 square meters. The model is updated every 15 minutes, enabling significant improvements in forecasting extreme weather events.

The startup's technology leverages massive datasets, including previously untapped sources, to gain an intricate understanding of the Earth's atmosphere and natural world phenomena. With over five petabytes of training data, Jua claims its model outperforms incumbents in accuracy, precision, and data exclusivity by orders of magnitude. Additionally, the deep learning-powered model is highly energy-efficient, requiring over a thousand times less computing power than other numerical weather models.

Jua aims to provide new levels of predictive ability to industries heavily impacted by climate change, such as energy, agriculture, transportation, and insurance. For the energy sector, the company's AI promises optimized power and demand forecasting, enabling more informed infrastructure and trading decisions, as well as unmatched early identification of major weather events that could impact energy grids.

In February 2024, Jua secured USD 16 million in seed funding led by 468 Capital and the Green Generation Fund to build its foundation AI model for the natural world. The company was co-founded in 2022 by serial entrepreneurs Andreas Brenner and Marvin Gabler, with hubs in Zurich, Berlin, and Cape Town.