Weather is notoriously unpredictable. Its effects, which are only growing more intense with climate change, can be devastating—physically, financially, and emotionally.
Causal Labs wants to change that. The AI startup, founded by ex-Cruise machine learning engineers Kelsie Zhao and Dar Mehta, has emerged from stealth with $6 million in funding led by Kindred Ventures, Fortune can exclusively report. Refactor, BoxGroup, Factorial, Otherwise, Karman Ventures, and angel investors also participated in the round.
At Causal Labs, Zhao and Mehta are developing an AI model that goes beyond current weather forecasting technology. “More than just predicting the weather, our model is interpreting the weather to directly recommend what is the optimal action you need to take to get to your intended and the desired objective,” Mehta said. “So, for example, instead of just telling a farmer it's going to rain, we can help a farmer determine the optimal time to harvest their crops or spray fertilizer in order to increase crop yields.”
The founders also want to do what traditional weather forecasting technology does, but better. Taking in data from satellites, their model provides updates in real time, whereas current weather prediction technology often takes upwards of six hours to generate updates, resulting in outdated weather forecasts, Zhao explained. Causal Labs’ model will also be more specific to users’ locations, rather than providing an average forecast across a large area.
These forecasts will be visible on a dashboard where users can zoom into their specific locations. For areas of the world without internet access, Causal Labs may need to provide its services a bit differently. “[W]hat most people don't realize,” said Mehta, “is that the granularity and the level of the frequency of updates is actually 10 times, or even 100 times, worse in other parts of the world compared to the U.S.”
As weather affects virtually all sectors, Causal Labs has been testing across a wide range of industries. The startup’s intended customers are businesses and government agencies that rely on weather forecasts to operate, like in agriculture, aviation, or energy. “The way we are partnering with these companies is by understanding what is the key decision that they’re making based on weather forecasts,” explained Mehta.
While many large language models process reams of text to generate human-like writing, Causal Labs’ AI model is trained on data such as satellite images. “[T]here's a very big distinction between what we're doing and LLMs as we know it,” said Zhao. Causal intelligence also has to meet high safety standards, given its real-world use, something that the cofounders are familiar with from their work at Cruise. At the core of Causal Labs’ technology is the how and the why. “[We’re] building a physics model that understands how the atmosphere is going to evolve, understands the cause and effect nature of the atmosphere and how there's inherent uncertainty in it,” said Mehta. Understanding how the weather works lets Causal Labs tell you what the weather is—which means you’re better prepared to deal with it.
See you tomorrow,
Nina Ajemian
Email: nina.ajemian@fortune.com
Submit a deal for the Term Sheet newsletter here.
Nina Ajemian curated the deals section of today’s newsletter. Subscribe here.
Update: After publication, Causal Labs informed Fortune that the final amount raised was $6 million, up from the $5.5 million it previously indicated.
