Model the domain
Specimens are structured around the observations a real beginner can make.
Rock Atlas
A polished geology explorer rebuilt from a class study tool into a public portfolio project with searchable specimen data, field notes, and transparent scoring.
0Specimens
0Minerals
0Rocks
0Field notes
Interactive app
Explore by observation
Search by trait, then layer in field observations. The ranked list updates instantly and keeps the uncertainty visible.
Reference workflow
Field Notes
Short, practical guides turn the data model into a field-ready way to think.
Data transparency
Reviewed reference data
The dataset is curated for educational identification practice and reviewed against public USGS and NPS geology material. It favors observable traits over exhaustive mineralogy.
Portfolio build
A static app with real product behavior
Rock Atlas demonstrates frontend architecture in a small surface: modular data, deterministic scoring, accessible controls, responsive layout, local persistence, and Docker-friendly deployment.
I am not a geologist. This is a portfolio app built with AI-driven development, public educational references, and personal curiosity; it should be treated as an educational tool, not professional geologic advice.
Keep uncertainty visible
Candidate scores are hints, not diagnoses, and every profile keeps common mix-ups nearby.
Ship the craft
The app stays dependency-free, fast, readable, and easy to deploy in the existing nginx stack.