Kd Bartholomew
Data Engineer @The Nature Conservancy
Signup · Get unlimited contacts
WORK HISTORY
Data Engineer @The Nature Conservancy
San Francisco, CA, US
Built a globally scalable, distributed surface water detection pipeline using NDWI, Otsu thresholding, and Canny edge detection, processing 3TB+ of satellite imagery weekly across 5 pilot sites. Deployed on AWS (S3, EC2, Lambda) with autoscalinginfrastructure and orchestrated via Apache Airflow.• Replaced physical stream gauges with an automated satellite-based system, delivering $225K in first-year savings across 5 test sites with projected exponential savings as coverage scales.• Developed XGBoost forecasting models with engineered seasonal and hydrological features to predict stream drying, boosting accuracy from 82% to 93% and outperforming SARIMAX baselines across multiple watersheds.
EDUCATION
University of San Francisco
Master of Science - MS, Data Science
University of San Francisco
Bachelor of Science - BS, Data Science, Cum Laude
ABOUT KD BARTHOLOMEW
I\'m a Data Scientist and Data/ML Engineer who builds end-to-end ML systems spanning ETL pipelines, data warehouses, model development and deployment, CI/CD pipelines, and dashboards. Proficient in Python, modern ML frameworks, and cloud infrastructure, I create scalable, production-ready solutions. I am passionate about modeling intelligence and building intelligent systems.In building systems that leverage massive amounts of data, I’m driven by the potential for proportionally large impact. It’s important to me to work in roles where changing the world for the better is a core goal. I’ve really enjoyed how pervasive that mindset is in conservation tech, nonprofit work, and research.I have a strong mathematics background, with academic research focused on stochastic processes, particularly Brownian motion, as a mathematical framework for probabilistic learning. I\'m interested in how these continuous-time models, especially Langevin dynamics, parallel modern machine learning algorithms that learn through noise, adaptation, and exploration. While Brownian motion describes learning as physical diffusion through uncertainty, ML systems approximate that behavior computationally, translating the mathematics of randomness into scalable algorithms for intelligence.
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.