Zurich-based startup Jua has secured $16 million in seed funding to build foundation AI model aimed at predicting and analyzing natural world phenomena like weather. The round was led by 468 Capital and the Green Generation Fund, with participation from Promus Ventures, Kadmos Capital, Flix Mobility founders, and others.
Jua is developing what it calls “foundational models for physics” - essentially a machine learning system that incorporates massive datasets to gain an intricate understanding of the physical world. The company’s first model focuses on highly accurate weather and climate forecasting, claiming capabilities like 16-day forecasts and "extreme high accuracy" surpassing traditional models.
The company says its model is fueled by tens of millions of primary data points from previously untapped sources, as well as petabytes of training data. It claims over 5 petabytes of training data—over 20x Google’s DeepMind GraphCast. This allows it to provide insights no other weather model can offer, outpacing incumbents in accuracy, precision, and data exclusivity.
Chris McKay
Ultimately, Jua aims to bring new levels of predictive ability to industries like energy, agriculture, transportation and insurance that need to respond to climate change dynamics. Volatility stemming from shifting weather patterns and climate impacts led over $100 billion in damages globally last year. Jua’s model seeks to arm these industries with better planning tools.
Specifically for the energy industry, Jua says its AI can enable optimized power and demand forecasting, empowering more informed infrastructure and trading decisions. It also promises unmatched early identification of major weather events like storms that could impact energy grids.
The company aims to eventually expand into areas like material sciences, biomedicine and chemistry. But its founders, CEO Andreas Brenner and CTO Marvin Gabler see immediate global demand from climate-exposed industries like agriculture, transportation, and insurance needing AI-powered insights to respond to mounting climate volatility.
Jua seems poised to join a new wave of European AI players focused on building homegrown models as alternatives to American AI giants like OpenAI, Google and Microsoft. Investor 468 Capital sees Jua’s model enabling industries worldwide to gain “powerful insights” about the natural world.
The startup however faces uphill challenges in continuing to amass training data and compute resources at scale. It will also need to convince potential customers its predictions are more accurate and valuable than existing weather models. Safety and transparency around such an influential model are also paramount.
Nonetheless, Jua's technology holds intriguing potential if it lives up to its promises. With climate volatility accelerating globally, there is no shortage of demand for AI-driven insights that this European upstart seeks to provide.