STEP 1
Research intelligence
Pre-field planning
What the published literature says about your terrane, your commodity, your deposit style — before a dollar is spent on the ground.
Technology
CAGE-IN. One engine, literature to drill targets.
CAGE-IN is AiRE’s flagship targeting platform — the engine proven on live programs and behind every result on this site. It isn’t a list of products; it’s one exploration workflow that takes a project from “what does the science say?” to “drill here, and here’s why.”
What CAGE-IN does
Four steps. One engine. One defensible verdict.
The same platform carries a project end to end — each step feeds the next, and every output traces back to cited science. The core is the multi-evidence scoring engine; the steps around it get your data in and your decisions out.
STEP 2
Regional reconnaissance
Satellite screening
Satellite-driven regional scanning — spectral screening over large areas to shortlist ground worth the full treatment.
STEP 3 · THE CORE
Multi-evidence AI scoring
The CAGE-IN engine
The flagship engine. Multi-sensor ingestion and spectral processing across six satellite platforms, GPU-accelerated — imagery, geophysics, geochemistry and drilling scored together by literature-calibrated deposit models, with ensemble fusion.
STEP 4
3D delivery
Outputs you can act on
Ranked target maps, sections, interactive single-file 3D deposit viewers your whole board can open, and prospectus-ready reporting.
Use cases
Wherever your project is, CAGE-IN plugs in.
From grassroots ground to a producing mine, the same engine solves a different problem at each stage of the exploration lifecycle.
Grassroots groundProducing mine
01 · Reconnaissance
Grassroots · any commodity
Find the ground worth chasing.
With only a few layers of data, CAGE-IN ranks where a mineral system is most likely to occur across a large area — so scarce budget goes to the ground that deserves it.
Minimal data in · any commodity02 · Targeting
Early-stage exploration
Drill targets in three dimensions.
From 3D ground geophysics, CAGE-IN places ranked drill targets in 3D — not just where to drill, but how deep, and why each one scores where it does.
~2× the conventional hit rate*03 · Optimisation
Active drill program
Fewer holes, better placed.
For a program already turning the rig, CAGE-IN re-scores against every new intercept to optimise where the next holes go — maximising return on every metre drilled.
Maximise ROI per metre04 · Resource expansion
Advanced projects & operating mines
Make the discovery bigger.
CAGE-IN reintegrates and reinterprets all existing data to find extensions and repeats around a known deposit — adding tonnes and extending life-of-mine.
Bigger discoveries · longer mine life*Typical of AiRE programs — prioritized targets drill-tested at roughly double the hit rate of conventional targeting. The comparison basis is available to qualified parties under NDA.
Inside the CAGE-IN engine
Three independent engines. One honest verdict.
CAGE-IN never trusts a single method. Two fundamentally different scoring engines run independently, and a third weighs their agreement — so methodological disagreement is surfaced to you, not averaged away.
A
Method A — Geological reasoning
Weighted evidence scoring through twelve deposit models calibrated to peer-reviewed literature, with a consistency veto filter that kills geologically incoherent scores.
B
Method B — Machine learning
A supervised classifier trained with spatial-block cross-validation — the discipline that stops spatial autocorrelation from flattering the accuracy numbers.
A+B
Ensemble fusion
Dempster-Shafer evidence theory merges the two, surfacing high-confidence agreement zones and flagging where the methods disagree — disagreement is information.
Open the full method detail — evidence layers, deposit models, citation trace
Every run fuses a multi-evidence stack — remote sensing, geophysics, geochemistry, geology and structure, terrain and drilling — and each layer enters the deposit model with its own model-specific, citation-locked weight, then de-conflicts into one ranked verdict. Whatever your project doesn’t have is disclosed as a dark layer, never quietly ignored.
Deposit models — literature-calibrated, citation-locked
Porphyry Cu-Au-MoEpithermal Au-AgIOCGSkarnIntrusion-related AuLaterite NiNi sulphideSEDEX polymetallicOrogenic AuMetamorphic graphiteVMSKimberlite (diamond)
One worked trace
Two contributing parameters from a literature-calibrated model, with their published sources:
| Parameter | Weight | Cited source |
|---|---|---|
| Ferric-iron alteration index | contributing | Sabins (1999) |
| Clay / sericite (SWIR) ratio | contributing | Crosta & Moore (1989); Hunt (1977) |
Engineering facts
220+ regression tests run on every release2-5 min per 100 km² processedGPU-adaptive tiling with CPU fallbackDesktop-deployed — your data stays on your machinesPer-run sidecar — inputs, versions & parameters recorded
Platform maturity
What’s proven, what’s expanding.
CAGE-IN is one platform with several capabilities. We label each one’s maturity honestly — the same discipline we apply to your geology.
| Capability | What it does | Status |
|---|---|---|
| Multi-evidence AI targeting | The core CAGE-IN engine — unified multi-sensor imagery + multi-evidence scoring | Operational on live client programs |
| 3D prospectivity & deposit visualisation | Ranked maps, sections and interactive single-file 3D deposit viewers | Production |
| Legacy data ingestion | Legacy archives → working GIS project, fast | Production |
| Regional satellite reconnaissance | Large-area spectral screening to shortlist ground | On the roadmap — expanding |
| Pre-field research intelligence | Literature-driven planning before fieldwork begins | On the roadmap — expanding |
See CAGE-IN on your ground.
The Prospect Brief tells you which capabilities your data can feed today — and what it would take to light up the rest.
Get the 48-Hour Prospect Brief