Case studies
Real AI. Real outcomes.
We don't stop at proof of concept. Every system we build runs in production and is measured by business impact.
ROM Blending Optimisation
ROI on blending optimisation
Challenge
Reduce ore grade variability and improve blend quality against grade and tonnage targets across a large manganese mining operation.
Solution
AI blending recommender that considers grade targets, ore stockpile constraints, and operational parameters to continuously recommend optimal blend configurations.
Milling Optimisation
Gold recovery improvement
Challenge
Improve gold recovery across ore silo filling, milling, and carbon-in-pulp processes at a large gold operation.
Solution
AA/AI models to homogenise ore quality via silo filling; optimised milling parameters including feed rates and densities; CIP modelling for carbon usage, activation, and fines management.
Predictive Maintenance
OEE improvement
Challenge
Reduce unplanned downtime and extend equipment life across a diamond mining operation's full asset fleet.
Solution
AI-driven predictive maintenance strategy; sensor specifications by equipment type; supporting data architecture; MVPs for select assets; work management framework integrating maintenance AI.
Energy Reduction
Energy consumption reduction
Challenge
Reduce energy consumption in a high-intensity nickel smelting operation where even small efficiency gains translate to significant cost savings.
Solution
Variability reduction models and process optimisation, delivering continuous real-time recommendations to operators running the smelter.
Schedule & Cost Recovery
Schedule delay recovered
Challenge
A greenfield mining construction project was 1 year delayed and 30% over budget. The client needed tools to regain control of schedule and cost.
Solution
Digital daily task scheduling application with real-time field feedback on progress and blockers; digital twin for the tailings facility with automated construction progress management.
Every engagement follows our 4-stage methodology
From strategy and ideation through MVP, production deployment, and ongoing support — with Go/No-Go checkpoints at each stage.
See how we work →