I am a geospatial data scientist based in the San Francisco Bay Area. My 25 years working with data and machine learning span astrophysics (2000-2012), generalist data science (2012-2025), and now geospatial (2025+).
Main areas of experience
- End-to-end solution of business problems for customers
- Problem definition
- Data preparation
- Modeling
- Deployment
- Results
- Machine learning: LLMs, deep learning, gradient-boosted trees, random forest, nearest neighbor, clustering, etc.
- Large datasets
- Communication of results and interpretations to all audiences
- Refereed publications
Tools
- Python: GeoPandas, PyTorch, TorchGeo, etc.
- QGIS / PyQGIS
- GDAL+OGR
- PostGIS / SQL
+ Many others
Recent Projects
These are some recent projects that are public
- Retrieval-Augmented Generation on Paperspace + DigitalOcean: Help Your Models Give Their Best Answers
- LLMs on DO+PS Multinode H100s: Pretraining and Finetuning MosaicML Models
- End-to-end Data Science on Gradient: Nvidia Merlin
For a portfolio of personal geospatial projects as part of my transition to this field, see here.
Next
For more, see work highlights, hire me / freelance, or the other links in the navigation bar above.