AI-driven Gold Prospectivity Mapping at Committee Bay

Auryn Resources faced a common exploration challenge at their Committee Bay project in Nunavut: identifying new high-grade gold targets across a massive 300 km strike length, much of which is masked by thick glacial-fluvial cover and shallow lakes. Traditional geochemical sampling is often impossible in these conditions. CGI stepped in with our proprietary VNet segmentation deep learning algorithm to integrate a vast array of datasets—including airborne magnetics, electromagnetics, and structural geology—to predict mineralization patterns.

By training the VNet on existing drill hole data until it could predict 99% of known mineralization, CGI successfully generated twelve new targets. These included critical extensions of the Three Bluffs deposit and an entirely new 15-kilometer-long structural corridor. The technology removed human bias and analyzed data sets beyond human cognitive capacity, giving Auryn the confidence to move these targets directly to the drill stage. Subsequent drilling validated the approach, intercepting 6g/t Au exactly where the VNet had predicted mineralization.