Featured Projects
Ore Exploration
Built a reproducible geoscience ML pipeline integrating geology, geophysics, and geochemistry to map critical mineral prospectivity. Combined Random Forests and Bayesian models with spatial cross-validation to identify top-ranked exploration targets, deployed as heatmaps and an interactive Streamlit app.
🔗 GitHub Repo · 📄 PDF Summary (in work) 🚀 Interactive Dashboard
ThermoML Parser
Parsed 1,200+ XML files to extract structured thermophysical data for ML-ready analysis. Designed CLI tools and uncertainty handling using modern Python. ThermoML-FAIR transforms decades of peer-reviewed thermophysical research into machine learning-ready data, making reproducible, sustainable materials discovery possible at scale.
🔗 GitHub Repo · 📄 PDF Summary
Thermal Conductivity ML Model
Used matminer to featurize materials from multiple datasets. Trained ensemble models achieving R² > 0.85 with SHAP-based interpretability. Now experimenting with deep learning + Docker deployment on a dedicated dev branch.
🔗 GitHub Repo 🔗 Dev Branch Repo ·
Selected Publications
I've co-authored peer-reviewed publications in Scientific Reports, ASHRAE Journal, Journal of Applied Microbiology, and SAMPE Conference Proceedings. These span aerosol modeling, materials testing, and surface engineering.
About Me
I'm Angela Davis, a materials scientist and data professional with 13+ years of experience turning complex engineering challenges into data-driven solutions. My work bridges materials R&D, machine learning, and sustainability, from advancing next-generation composites and coatings to building FAIR data pipelines for thermal conductivity and informatics projects.
I'm passionate about using AI and materials science to accelerate sustainable innovation, whether that's reducing waste, enabling cleaner manufacturing, or developing open-source tools for the scientific community.
Beyond my technical work, I thrive on collaboration, mentoring, and storytelling with data—helping teams connect insights to impact.