Living Methodology
Most exploration software is frozen at the day it shipped. Ours reads the literature.
Ore-deposit science moves monthly — new geochronology, new alteration vectors, new exploration criteria, published and peer-reviewed. AiRE’s deposit models are maintained as living, versioned scientific documents that absorb that progress. When you score your ground with AiRE, you score it against today’s understanding — not the understanding from whenever your software vendor last released.
How a model stays alive
From journal page to scoring engine — with a paper trail.
The literature watch
The council's literature seat monitors peer-reviewed publication across economic geology — journals first, then quality-assessed gray literature from the major geological surveys. Relevance and credibility are scored separately; predatory sources are excluded outright.
Dual domain review
A proposed model change is reviewed twice — once for the geology (does this criterion genuinely discriminate?), once for the citation (does the source actually support this weight?). Either reviewer can veto.
Ratification & locking
Accepted changes are versioned and cryptographically hash-locked, with the previous weights preserved in the audit trail. A model can always answer: what changed, when, on whose evidence?
Regression-guarded release
Every model update passes the full automated test wall — 220+ regression tests — before it can touch a client run, so every release reproduces the same answer on the same data. Science evolves; rigour doesn't relax.
The honesty architecture
Four promises every AiRE deliverable keeps.
No weight without citation
Every parameter in every deposit model carries its published source. If the science can't be cited, the weight doesn't ship — there are no folklore numbers in the engine.
Coverage disclosure, every run
Each score states what fraction of the deposit model's evidence your data could actually test — and names the dark layers with reasons: no input data, capability gap, or data-quality reject. A high score on thin evidence is flagged as exactly that.
Consistency vetoes
A criterion-based veto filter kills geologically incoherent scores before they reach your map. High numbers that violate the deposit model's own logic don't survive.
Reproducibility on demand
Every run emits a machine-readable record of its inputs, parameters, model versions and software state. Six months later, your QP can ask “how exactly was this produced?” — and get a complete answer.
Evidence discipline
When sources disagree, rank them — don’t average them.
AiRE operates a written authority ranking for conflicting evidence: direct field observation outranks modeled surfaces; peer-reviewed literature outranks company reports; measured geometry outranks statistical inference. Conflicts are resolved by rank and recorded — never silently blended.
Score your ground against today’s science.
And re-score it next year, when the science has moved — that’s the point.
Get the 48-Hour Prospect Brief