
Tomorrow.io extends Series F to $210M with $35M new funding for weather forecasting
The AMW Read
Incremental funding update to an existing player; segment-level significance as a vertical AI + infrastructure company it carries segment-level significance for AI-agents and data-moat dynamics, but does not resolve an open debate.
Tomorrow.io extends Series F to $210M with $35M new funding for weather forecasting
Tomorrow.io, the US-Israeli weather intelligence unicorn formerly known as ClimaCell, announced an additional $35 million in Series F funding from Pitango, Harel Group, HarbourVest, and Stonecourt, bringing the round total to $210 million. The company, founded in 2016 by IDF veterans Shimon Elkabetz, Rei Goffer, and Itai Zlotnik, has now raised approximately $535 million in total capital.
The fresh capital is earmarked for deploying DeepSky, a commercial satellite constellation purpose-built for the AI era, and for expanding the company's agentic AI platform that delivers real-time operational intelligence. Tomorrow.io positions weather data as an under-integrated force in global economic decision-making, now becoming foundational as enterprises embed AI into operations. The company employs over 150 people and maintains a small presence in Israel.
Why it matters: Tomorrow.io exemplifies the "vertical AI + proprietary data moat" pattern β a startup that captures a high-value physical-world domain (weather) by combining first-party sensor infrastructure (satellites) with agentic AI inference. This is a structural play against hyperscaler commoditization: the company does not rely on third-party APIs from large foundation-model labs; it owns the observation layer and the decision layer. The $210M Series F (well above typical late-stage rounds) signals that institutional investors see weather intelligence as a defensible AI category, not a feature of general-purpose forecasting. The bear case remains whether a 150-person team can sustain a satellite constellation against NOAA/ECMWF public investments and Google DeepMind's GraphCast.
