I build production-grade automated pipelines that bridge the gap between advanced Computer Vision models and multi-spectral geospatial data. My work focuses on scalable raster processing, automated land-cover classification, and spatial engineering.
- Core & Architecture: Python, Flask, FastAPI, Databricks, PySpark
- Geospatial Engineering: Rasterio, NumPy, Shapely, Fiona, GDAL/OGR, PyProj
- Computer Vision & ML: PyTorch, Segment Anything Model (SAM/SAM2), OpenCV, Scikit-Learn
- DevOps & MLOps: GitHub Actions (CI/CD), Docker, Poetry, Pytest
- Large-scale Rasterization & Vectorization: Designing zero-loss geometric conversion pipelines.
- Top-Down Computer Vision: Adapting foundation models (like SAM) to overcome scale-invariance and spatial coordinate alignment in satellite/aerial imagery.
- Automated Data Orchestration: Building robust, time-triggered workflows for environmental and green-space health analytics.
