Technology & Methodology
Implementation
Project focused on data validation, interpretation, and visualization that analyzes critical insights through open-access technology, fostering the democratization of information.
Data Recollection
Community-Led Data Collection with KoboToolbox
A project that utilizes KoboToolbox, a premier open-source platform for field data collection, to capture real-time socioeconomic and environmental insights across Latin America. By deploying responsive and offline-ready digital surveys, the initiative gathers critical ground-truth data on food security, water access, and climate impact. This approach replaces traditional paper-based methods with a high-integrity digital workflow, ensuring that community voices are accurately represented and seamlessly integrated into the project’s analytical ecosystem.
Data Validation
Data Integrity and Metadata Standardization
A project that leverages Open Data Editor (ODE), an open-source tool developed by Frictionless Data, to guarantee high-quality and interoperable datasets. By integrating ODE into the data architecture, the project streamlines the validation of data integrity and the automated generation of metadata, ensuring that all collected variables are standardized and robust. This implementation allows for a rigorous quality control process, enabling the seamless transformation of raw data into reliable insights for public policy simulation.
Predictive Simulation Modeling
Predictive Modeling with XGBoost Simulations
A project that implements an advanced XGBoost regressor to simulate complex health scenarios and predict outcomes such as low birth weight (BPN) based on environmental and socioeconomic factors. By training on high-dimensional datasets, the model identifies non-linear relationships between variables like air quality (NO2), poverty levels, and health infrastructure. The resulting prediction datasets are integrated into three-dimensional data cubes and visualized through interactive Looker Studio dashboards. This allows decision-makers to manipulate input variables and observe real-time simulations of how specific policy changes or climate events could impact public health across the region.
Data Visualization
Interactive Intelligence: Multi-Country Data Visualization
A project that utilizes Looker Studio to transform complex, multi-dimensional data from Argentina, Chile, Colombia, Uruguay, Peru, and Paraguay into interactive, high-impact visual intelligence. This fundamental component of the ecosystem integrates a comprehensive data cube—combining remote sensing, health indicators, political variables, and environmental metrics—to reveal critical insights hidden within the datasets. Beyond descriptive analysis, the dashboards feature dedicated simulation modules powered by an XGBoost model. These modules allow users to interact with predictive scenarios, visualizing how modifications in environmental or political factors directly affect public health outcomes over time.
How data was processed?
Data Collection
A project that conducts massive searches for open-access data from government portals and national statistical institutes. This stage involves gathering political, sociodemographic, and health indicators across participating countries, ensuring a comprehensive regional dataset for multi-scale analysis
Satellite Intelligence
A project that extracts environmental and climate data from open portals such as Worldclim and Google Earth Engine. By processing images from missions like Sentinel-2 and Landsat 8, the system calculates critical environmental indices, including air quality, water safety, and heat waves.
Validation & Democratization
A project that utilizes the Open Data Editor to guarantee data integrity and standardized metadata generation. By publishing validated datasets on repositories like Zenodo, the initiative ensures the democratization of information, providing transparent and reproducible scientific resources for the global community
Data Cube Architecture
A project that constructs three-dimensional data cubes within BigQuery to integrate collected metrics and simulation results. These structures enable high-performance analysis and allow the system to efficiently process complex relationships between environmental, social, and health variables for seamless consumption in visualization tools.
Decision-Making Dashboards
A project that delivers dynamic dashboards in Looker Studio, providing decision-makers with a simple way to visualize health impacts. These interactive tools facilitate evidence-based public policies by allowing real-time exploration of climate scenarios and predictive policy interventions.
Ethical Governance & Data Integrity
Data Stewardship: FAIR & CARE
A project that aligns its data management with the FAIR (Findable, Accessible, Interoperable, Reusable) and CARE (Collective Benefit, Authority to Control, Responsibility, Ethics) principles. This dual approach ensures that technical excellence in data sharing is balanced with the ethical protection of community interests and the sovereignty of regional information. By implementing these standards, the initiative guarantees that all processed data serves as a transparent, high-quality, and culturally responsible resource for Latin American public health.
Key Data Science Principles:
- Privacy
- Autonomy
- Transparency
- Equity
- Solidarity
- Responsibility
- Integrity
- Quality
- Security