The people behind Yakana's Team

Led the project end to end: set the vision, coordinated the team and the work of every branch, and made sure everything came together as one coherent product. Designed the visual identity and shared design system, built the Story, Prevention, Response and Team pages, the EN/ES language system and the connection to the real data API, automated the daily data refresh and the deployment, and kept the quality of the whole site high.

Built the Method page: the before-and-after chart, the validation and the interactive simulator of the harmonization formula. Also designed the official Yakana logotype and isotype.

Built the Explorer: the department map, the 3D terrain view with real hotspot cells and the 23-year monthly calendar. Also built YakaLlama, the assistant connected to NVIDIA models and the backend data, and produced the project's pitch video.

Built the API that serves the historical series, the live week and the prevention model, and configured and deployed it on Google Cloud so the whole platform can query it.

Engineered the data ingestion pipeline, implemented the MODIS–VIIRS harmonization methodology, ensured data quality, set up the monthly retraining of the model and developed the backend API powering the entire platform.

Built the predictive model, a logistic regression trained on the harmonized series and compared against four other methods, and collaborated on the MODIS–VIIRS harmonization model and on integrating the team’s work into the 2003–2026 series and its API.