Tool
XRF Data App
A browser app that reconciles handheld-XRF field readings against a master tracker — linking thousands of readings to samples for browsing, QC, and export.
The problem
A boulder-sampling campaign generated XRF readings from two different guns, each exporting its own CSV. Those readings had to be reconciled against a master tracker spreadsheet — matched by gun serial number and reading number — before anyone could actually use the data. Done by hand across thousands of readings, that’s both slow and a reliable source of mistakes.
What I built
A static web app that ingests the raw gun CSVs and the master tracker and gives the team one queryable interface over all of it:
- a drill-down browser (areas → grids → boulders → individual readings);
- a boulder detail view with tabs for surface XRF, depth XRF, powder/lab data, and QC;
- a customizable element picker (8 common elements by default, expandable to 45);
- reconciliation of unlinked readings, plus CSV export by reading range or sample ID.
How it works
Reading-to-boulder linking uses the tracker’s serial-number + reading-number columns as the canonical mapping, with CSV fallback validation to catch readings that don’t line up — so QC problems surface immediately instead of silently flowing into the dataset.
Outcome
Field data that used to live in disconnected spreadsheets became a single, browsable, exportable record — with the linking and QC handled automatically.