Public Health Organization

gravity9 designed and delivered a semantic enrichment framework that transforms how clinicians and analysts interact with openEHR data. Using MongoDB, vector embeddings, and large language models, we created a metadata template system, an automated enrichment engine, and a natural-language question answering workflow. Together, these components demonstrated how AI-driven enrichment can extract new meaning from clinical records, simplify data exploration, and enable users to ask questions conversationally no query language required. In just 13 days, gravity9 produced a working proof of concept that turned complex medical information into clear, actionable insights.

Our approach began by mapping current user interactions with openEHR and identifying common points of friction. From there, we developed a flexible metadata model and an enrichment pipeline capable of generating derived values, logic-based insights, and vector embeddings for similarity search. We then built a prototype workflow that translated natural language into precise database queries. Despite limited time and no real-world data, gravity9 successfully delivered all core components, proving how AI and MongoDB can modernize access to healthcare data and lay the groundwork for future clinical integration.