NASA democratizes Earth Science data with conversational AI agent

NASA operated satellites generating over 100 petabytes of Earth Science data faced major accessibility and usability challenges for both specialists and the broader public. Traditionally, navigating geospatial datasets required specialist knowledge and intense manual effort. NASA's Office of the Chief Science Data Officer partnered with Microsoft to develop Earth Copilot, an AI conversational assistant built using Azure OpenAI Service and hosted on Azure Cloud. Earth Copilot integrates with NASA's VEDA platform, enabling plain language queries to analyze complex geospatial data—such as querying hurricane impacts or evaluating air quality trends. The agent leverages scalable AI and machine learning capabilities of Azure to handle sophisticated queries on vast geospatial datasets. This solution has opened up direct access for researchers, educators, policymakers, and the public, who can now analyze NASA Earth data with ease and efficiency. Democratizing this access supports NASA's open science mission, accelerates scientific insight, and enables faster decision making in climate, disaster response, urban planning, and agricultural applications. The project has completed its proof of concept and is undergoing testing within NASA before broad public release. Data privacy, misuse prevention, and responsible deployment of AI remain core operational considerations.