LGND AI, Inc., a startup reimagining how people and AI engage with Earth data, has raised $9 million in a new funding round led by Javelin Venture Partners. The round also included participation from AENU, Space Capital, Overture, Ridgeline, MCJ, Coalition Operators, Clocktower Ventures, and several angel investors such as John Hanke (founder of Keyhole, the basis for Google Maps), Karim Atiyeh (cofounder and CTO of Ramp), and Suzanne DiBianca (EVP & Chief Impact Officer at Salesforce). Javelin’s Noah Doyle—formerly head of enterprise products for Google Earth and Maps—will join LGND’s board.

LGND aims to make Earth data more usable and accessible by building infrastructure that transforms complex geospatial information into intuitive, AI-friendly formats. Traditionally, Earth data has been underutilized in AI workflows due to its high cost and complexity. LGND addresses this by creating dynamic “geographic embeddings”—a next-gen data layer that allows AI models and human users to interact with geospatial data more intelligently.

Instead of relying on conventional pixel-based mapping or static datasets, LGND has developed a “geo-embeddings factory” capable of producing scalable, queryable, and adaptable Earth data assets. These assets can evolve in real time, becoming more accurate and insightful the more they’re used. The company likens this innovation to how map tiles once revolutionized digital mapping—only now with a deeper, more context-rich layer.

CEO and cofounder Nathaniel Manning said the company’s mission is to make Earth data universally actionable through AI. By embedding geospatial awareness directly into AI systems, LGND envisions broad impact across sectors such as insurance, defense, logistics, and climate adaptation.

So far, LGND has seen early traction through pilots in areas like wildfire risk modeling, illegal mining detection, and infrastructure monitoring. The company’s tools are already being explored by professional service firms and logistics companies to integrate geospatial intelligence directly into operational and decision-making workflows.