Stratum AI applies deep-tech machine learning to the mining industry, positioning itself as the world's first AI mineral resource modeling company. The company's work sits at the intersection of geostatistics, deep learning, and unconventional data processing, aiming to improve material extraction while reducing environmental impact. Its headquarters are in Toronto, Ontario.
The firm's primary product is SAIGE (Stratum AI Geostatistical Engine), a platform that uses deep learning to produce mineral resource, ore reserve, and geometallurgical models. It is designed to predict the underground distribution of minerals such as gold, silver, and copper, and its developers claim it yields models significantly more accurate than traditional Kriging methods.
Stratum AI has reported specific operational improvements from deploying its technology. These include a 41% reduction in waste produced, a 6% gain in yield, a 58% reduction in tonnage deviation, and a 20% increase in high-grade stockpile grade at client operations. The system has also been associated with approximately 25% fewer drilling requirements.
