Tomorrow.io Announces DeepSky, the World’s First AI-Native Space-Based Weather-Sensing Constellation

Tomorrow.io, shaping the future of global resilience, led by Shimon Elkabetz, Rei Goffer, Itai Zlotnik, Leigha Kemmett, Osnat Barak, and others, has announced DeepSky, the world’s first AI-native, space-based atmospheric and oceanic sensing network, designed to make Earth’s atmosphere and oceans continuously observable in real time, powering faster, smarter global decision-making and response.

This marks Tomorrow.io’s second satellite constellation, announced shortly after completing full deployment of its first constellation featuring Ka-band radar and microwave sounders.

DeepSky positions Tomorrow.io to lead a new generation of AI-native, resilient, and commercially scalable weather infrastructure. As AI improves forecasting accuracy and efficiency, limitations are increasingly driven by gaps in global observation coverage rather than models or compute, highlighting the need for denser, high-frequency, and diverse satellite data.

DeepSky is designed to address gaps in global weather observations by delivering the temporal density and sensing diversity required for modern forecasting. The proliferated LEO constellation uses highly capable satellites with proprietary instruments to provide higher revisit rates, faster model refresh cycles, and improved extreme weather prediction. It complements existing government GEO and LEO systems by extending coverage and enabling new sensing capabilities.

DeepSky’s design centers on multi-modal sensing across the electromagnetic spectrum, high-revisit global coverage to increase temporal density, and new sensor classes previously limited to specialized science missions due to cost or revisit constraints.

DeepSky combines scale and advanced sensing to deliver observation cadences that were previously impractical, while reducing cost per scan and impact. It is built to serve a broad range of operational users, from meteorological agencies and forecasting centers to defense and international partners, enabling more adaptive, AI-driven decision systems. The constellation is designed to exceed baseline observational requirements and remain flexible for evolving missions and applications.

DeepSky delivers frequent, diverse, and globally consistent observations to enable faster model refresh cycles, improved extreme weather prediction, and new AI-native applications not possible with current observation density.

Modern supply chains increasingly require continuous sensing and real-time intelligence to drive network-wide decisions. DeepSky advances this shift by enabling a unified, AI-driven operational picture, building on Tomorrow.io’s experience in commercial weather satellites and global data delivery. The constellation will be deployed in phases, with further details released as development progresses.