• We aim to operate the world's most intelligent mines with

    Artificial Intelligence

  • Creating the Resource Model

    • 1. Collect Drill Hole Data

      Data is collected by physically drilling 100m into the ground, and then removing each rod

    • 2. Analyze Core Samples

      Samples are taken at multiple depths of each rod which are sent offsite to a chemical lab to get assayed. The assays on the core samples measure the mineral grade.

    • Mill Data
      Blasthole Data
      Assay Data

      3. Collect Blasthole + Mill Data

      Stratum's model learns from higher density data, such as production data, to model lower density data, such as drillholes.

    • 4. AI Resource Model Construction

      The Stratum AI model learns geological patterns from all historical data to construct more accurate mine resource model

  • The Stratum Resource Model

    Explore an example of our AI model below.

    Depth Profiles

    Model identifies high grade lode previously undetected by kriging

    AI more accurately determines boundaries of main mineralization

    AI extrapolates from several known samples the direction of a vein

    Depth Profiles

    Click below to check out cross section profiles at different depths.

  • How We Compare

    • Better mine planning + tonnage forecasting: 8-12% more resource mined
    • Better waste classification: 20-40% reduction in waste sent to mill
    • Better mill blending: optimized for mine economics and sales contracts
  • Contact Us


    Daniel Mogilny


    Farzi Yusufali


    Ilia Sucholutsky

    VP Research

    David First

    Chief Geologist