DomainMCF

DomainMCF (Maptek Compute Framework) is a cloud-based machine learning tool for geological domain modelling, launched in 2021. It represents a fundamentally different approach: geologists feed in drillhole or sampling data, and DomainMCF uses ML to generate domain boundaries and grade models in minutes rather than the days or weeks traditional methods require. Processing runs on Maptek's Compute Framework cloud infrastructure.

DomainMCF is designed as an add-on working through GeologyCore. It is particularly suited to large-volume datasets from later-stage exploration and operating mines. The tool produces domain models that include uncertainty measures, enabling geologists to identify where the ML model is less confident. Results download as block models publishable to Vulcan for downstream grade estimation. Validated through real-world deployments, including Roy Hill (iron ore) where domain flagging took under 50 minutes versus weeks using traditional methods.

DomainMCF's differentiators include its cloud-native subscription architecture, built-in data validation, domain uncertainty quantification, and rapid regeneration capability as new drilling data arrives. Since 2023, it offers persistent machine files for auditability, flexible block sizing from single model runs, and sample point distance measures. DomainMCF is positioned as a complement to — not replacement for — conventional modelling techniques.