STRATOS: Spatio-temporal Text-to-SQL for meteorological data

The growing availability of massive data from Earth observation programs such as Copernicus has opened unprecedented opportunities in sectors like precision agriculture, logistics, and renewable energy. However, its practical use encounters a technical barrier: the need to master query languages like SQL and specialized formats (NetCDF, GRIB) that are beyond the reach of business analysts or decision-makers. This is where the evolution of Text-to-SQL systems becomes meaningful, but applied to the spatio-temporal domain. The challenge is not trivial: there is a deep gap between symbolic concepts ('Mediterranean coastal area') and their numerical representation in coordinates and time scales. Solutions like the STRATOS framework (Spatio-Temporal Resolution Agent for Text-to-SQL) precisely address this leap, resolving semantic and spatial ambiguities before generating the query. From a business perspective, this capability allows any professional to ask questions in natural language —such as 'What was the average temperature in northern Spain during the last quarter?'— and obtain accurate answers in seconds, without the intervention of data engineers. At Q2BSTUDIO, we develop AI for businesses that integrates intelligent agents capable of understanding geographic and temporal contexts, transforming raw data into actionable insight. Additionally, our custom software solutions allow adapting these query engines to proprietary datasets, combining them with AWS and Azure cloud services to scale on demand. The key lies in combining natural language processing techniques with local ontologies, as state-of-the-art systems do, so that artificial intelligence not only translates words but also understands coordinates, time zones, and units of measurement. And this, in turn, can feed Power BI dashboards or automated reports, closing the loop between complex data and strategic decisions. For organizations looking to leverage meteorological information without investing in their own infrastructure, we offer AWS and Azure cloud services that ensure performance and cybersecurity at every stage of the flow. Ultimately, the convergence between natural language processing and spatio-temporal data is maturing, and companies that adopt these approaches —whether through custom applications or AI agents— will gain a real competitive advantage in a world where climate and location remain critical factors.