Skip to main content

Technology & AI

GeoAI4EI combines state-of-the-art geospatial and epidemiological modelling with advanced AI to turn multi-source data into actionable epidemic intelligence.

Foundations: the MOOD platform​

GeoAI4EI builds on the Horizon 2020 MOOD platform, an established foundation for geospatial and epidemiological modelling. This heritage provides:

  • Proven methods for spatio-temporal disease modelling
  • A track record of integration with public health data
  • A base to extend with modern AI techniques and new data sources

Geospatial & epidemiological modelling​

The core of the toolbox is a modelling layer that:

  • Represents disease processes in time and space, at national, regional, and local scales
  • Supports nowcasting, forecasting, and scenario analysis for epidemic preparedness
  • Combines epidemiological knowledge with data-driven methods to improve robustness
  • Produces uncertainty-aware outputs, so decision-makers can see the confidence behind every estimate

Multi-source data integration​

GeoAI4EI fuses data from six complementary domains — epidemiological, mobility, environmental, ecological, Earth Observation, and socio-economic — to build a richer picture of where, when, and how disease risk emerges. See Data integration for details.

AI: LLMs & retrieval-augmented assistants​

Beyond classical machine learning, GeoAI4EI integrates:

  • Large Language Models (LLMs) for summarising evidence, answering domain questions, and assisting with analysis
  • Retrieval-augmented generative (RAG) conversational assistants that ground answers in the curated, governed data corpus — reducing hallucinations and keeping answers auditable
  • Explainability and transparency measures, including model cards and uncertainty visualisation
caution

Conversational assistants are decision support tools. GeoAI4EI designs safeguards against over-reliance: every answer is traceable to its sources, and no output replaces professional public health judgement.

Platform design​

  • Open-source: all tools are published openly for reuse and inspection
  • Lightweight: designed to run in resource-limited contexts
  • Interoperable: aligned with EU infrastructures (EHDS, ATHINA)

Security & hosting​

  • EU-based secure hosting for all project data and services
  • Encryption in transit and at rest
  • Role-based access control to enforce the principle of least privilege

See Ethics & Governance for the full ethical and legal framework.