I've spent over 14 years building geospatial data infrastructure across government, environmental science, and enterprise platforms, from national SDG indicator pipelines at Chile's Ministry of Environment to educational GIS systems for Venezuela's Ministry of Education (FUNDABIT). My M.Sc. thesis (Universidad Mayor) analyzed post-fire structural change in Pinus pinaster forests using airborne and terrestrial LiDAR.
That remote sensing foundation has led me into computer vision: a zero-shot water body detection pipeline using Meta's SAM3 on 705 Landsat scenes of Chile's Laguna de Aculeo (IoU 0.95 against a calibrated statistical baseline), a YOLOv8m + SAM3 building and roof detection pipeline evaluated across ~98,000 tiles in Santiago, and a Mask R-CNN model for building footprint extraction validated on Spain's national cadastral data. I also built and published an open InSAR deformation-mapping pipeline used during the 2026 Venezuela earthquake response.
Most GIS and remote sensing tooling that reaches this level of rigor sits behind enterprise licenses that most practitioners in Latin America can't access. Sponsorship funds the freemium tools I'm building instead: a growing set of open source apps and Python packages for geospatial analysis, priced (or free) for individual practitioners rather than institutions. It also supports documentation and technical resources in Spanish, so the reasoning behind these methods, not just the code, is accessible to the regional GIS and remote sensing community.