Exploration Geochemistry & Data Science Consultancy
I help exploration teams get more from their geochemical and mineralogical data. With 15 years across mineral exploration, quantitative mineralogy, and applied machine learning, I build reproducible workflows that turn raw assay and geoscientific data into interpretable, decision-ready outputs — from QA/QC, preprocessing, and levelling through to multivariate analysis, predictive modelling, and interactive reporting.
All work is delivered as interactive HTML reports your team can explore, filter, and discuss in any browser — no specialist software required.
Services
QA/QC, Preprocessing & Levelling
Reliable interpretation starts with reliable data. I QA/QC, clean, preprocess and level multi-element geochemical datasets so that downstream analysis is built on a solid foundation.
Even on its own, properly preprocessed data sharpens anomaly detection and produces more reliable targets. It also dramatically improves outputs from any downstream multivariate analysis.
Unsupervised Multivariate Analysis
Multivariate geochemical data contains hidden patterns — alteration trends, geochemical domains (e.g. rock types, metallurgical domains), and element associations that reflect underlying geological processes.
Unsupervised methods extract these patterns without preconceived labels, giving exploration teams unbiased insights that can reinforce existing models or identify new targeting opportunities (e.g. alteration vectors).
Supervised Machine Learning
Supervised models trained on existing geological data can predict rock types, alteration styles, prospectivity, and material properties in new or unlogged samples — amplifying the value of work already done and allowing teams to test new or existing hypotheses.
Explainable ML and uncertainty quantification give geologists the tools to critically evaluate every prediction, rather than simply accept it.
Contact
- Email: burkettdane@gmail.com
- LinkedIn: Dane Burkett