Pre-Assay Geological Intelligence for Cu-Au Porphyry Systems.
Cauldra turns standard core logging data into a quantitative read on grade — predicting high-grade intervals and ranking the geological controls behind them, weeks before assays return from the lab.
Step-outs, hole extensions, target ranking, lab queue priority — the calls that shape a program happen while the rig is turning and the lab is weeks behind. Logging already captures what should predict grade. Cauldra makes it quantitative — before assays return.
When Cauldra deprioritises an interval, it is correct more than nine times in ten.
A ranked list of intervals by likelihood of hosting high-grade copper, with confidence tiers and geological reasoning. Project geologists adjust hole plans and step-outs weeks ahead of the lab.
Independently recovers the geological controls operators document — validated across three independent BC porphyry systems.
Cauldra identifies which lithologies, alteration assemblages, and sulphide associations matter — produced from logging alone, without access to the operator's interpretation.
The model's predicted probability that each logged interval exceeds the deposit's Cu p90 threshold. No assay results seen by the model — predictions made from logging alone. Hover any interval for depth, probability, lithology, and alteration.
Source: Publicly disclosed BC ARIS assessment report data. Hole identifiers anonymised for public display.
Cauldra has been trained and validated across three independent BC porphyry systems — different terranes, different operators, multiple drill seasons. Over 300 drillholes and 130,000+ metres of logged core. The case study below uses publicly disclosed BC ARIS drill data from a North American Cu-Au porphyry project across three consecutive programs (2021, 2022, 2023). Cauldra was trained on the 2021 and 2022 logging records and ran on the 2023 program blind. The model's outputs match the geological controls documented in the operator's published assessment reports — produced from logging alone, without access to the operator's interpretation.
The model recovers the porphyry zoning framework from logging alone — productive intrusive hosts, barren post-mineral dykes, sulphide paragenesis, and alteration zoning.
The geological controls on grade are learnable from standard core logging data — independently of the operator's interpretation. For an active program, that means a calibrated read on grade as logging comes in, weeks before assays return.
Trained and validated across three independent BC porphyry systems spanning different terranes, operators, and logging conventions.
| Validation type | AUC range | NPV |
|---|---|---|
| Same program | 0.75 – 0.80 | > 95% |
| Cross-year | 0.73 – 0.86 | > 90% |
| Cross-deposit | 0.72 – 0.82 | > 86% |
The model is not memorising a deposit. It is learning the porphyry signal that transfers between systems — with no retraining.
Strongly tied to logged sulphides and alteration. Validated across three independent BC porphyry systems with no retraining between deposits.
Nuggety and structurally controlled. Active work on where logging-plus-geochemistry can predict it reliably. Not claimed as current capability.
The geology shapes the model. The model sharpens the geology. That's the product.