Nature-Based Solutions (NbS) are interventions that leverage natural ecosystems to address societal challenges like climate change, public health, and biodiversity loss. Despite their growing implementation across Europe, data on NbS projects is fragmented across various platforms, making it difficult to analyze trends or evaluate impact.

As part of a broader initiative to centralize and standardize NbS data, I contributed to the development of a database that consolidates information from multiple sources to enable better monitoring, evaluation, and research.

Conducted a comprehensive survey of existing NbS databases and platforms, documenting their scope, data structures, and accessibility.

Automated the extraction of relevant project data (e.g. location, implementation dates, objectives, NbS types) using web scraping and Natural Language Processing techniques.

Developed and implemented a large language model (LLM)–based pipeline to identify and remove duplicate project entries across disparate data sources.

Parsed NbS objectives to extract and classify climate hazard targets (e.g., heatwaves, floods) as reported by implementers, linking them to potential health and well-being indicators.

Conducted descriptive statistics and contextual analysis using remote sensing data and LLMs to quantify trends, coverage, and environmental context of each NbS project.

Used the Una.city platform, a manually curated NbS dataset, as a validation benchmark for model performance and data quality assurance.