Leyline scans every land cell on Earth, then works region by region to find out which signals actually reveal ore on that ground, and which ones lie.
Leyline scans every land cell on Earth for the traces ore leaves behind. Public geoscience carries most of it: magnetics, gravity, lithology, structure, the shape of a drainage network. Some of it comes from places that have nothing to do with mining. Water chemistry. Soil gas. The way vegetation behaves over metal. Colonial survey ledgers written by people who walked the ground a century before anyone flew over it.
Then we narrow to one region at a time and work out which of those traces can be trusted on that particular ground.
Because the ground does not behave the same way twice. A signal that reliably marks gold under the cover of Western Australia can be worth nothing across the copper belts of central Africa. The rock is different, the weathering is different, the plants growing on top of it are different.
Almost every prospectivity model ignores this. One rule gets applied to the whole planet and the map comes out looking confident everywhere. We do the opposite. We test each signal against ground where the answer is already known, region by region, and we keep what fails as carefully as what works. A region we have worked through can be searched fast, with a reason attached to every answer.
Bedrock in the open. The alteration halo around ore is visible from orbit, and every operator in the industry can see it.
Closed canopy over the same kind of rock. The standard workflow masks this ground out before the analysis starts.
Satellite exploration reads alteration minerals on bare bedrock. It works, and it has one hard requirement: the rock has to sit in the pixel. Forest, regolith, sand and ice all defeat it. Whole basins get discarded before anyone looks at them, and exploration drifts toward the places that are easy to see rather than the places likely to hold metal.
We treat the cover as the instrument. Roots reach metres down and integrate the chemistry of a soil volume for decades. Streams carry evidence downhill from ground nobody can see. Termites build their mounds out of material excavated from well below the surface. None of that needs an outcrop.
Arid ground, read the rock. Covered ground, read what grows on it and what runs off it. Knowing which one to trust is the part that changes with the terrane.
Every method we test gets an entry. The entry names the region, carries the verdict, and keeps the evidence that produced it.
A verdict is local. The same method can hold on one terrane and fall apart on the next, so the verdict belongs to the pair, the method and the ground, never to the method on its own. Most systems decide once that a technique works, then apply it everywhere.
Failures stay in. Knowing that something has already been tried on your ground and returned nothing is worth as much as knowing what worked, and almost nobody publishes it.
A Full Scan draws on the codex for the ground it covers. Everything it learns goes back in.
The record grows with every region we work. Nothing we measure leaves it.
Four layers and you are still searching a province. Fifteen and you are standing on a licence.
Models that read further than any team can propose what to test. We test it ourselves, at continental scale, on ground where the answer is already known. Most candidates die there. The ones that live are ours.
Each region sharpens the next. Far enough down that road, finding a deposit stops being an expedition and becomes a lookup.
Licence areas, whole jurisdictions, or a question about a terrane we have already worked through. We reply to everything.
contact@leylineai.com