How We Work With Clients
6 months. 3 prediction cycles.
One continuously improving model.
4Point AI turns a mining company's existing exploration data into an iterative 6-month predictive targeting system. The first cycle builds the foundation. The second and third cycles incorporate new field evidence, reduce uncertainty, and refine where the client should spend its next drilling dollar.
6-Month Engagement Cadence
A repeatable cadence of predict · validate · refine.
Each prediction cycle delivers an updated 3D predictive model, ranked target list, and technical report. Between cycles, the client advances targets in the field. Every field result becomes input for the next update.
The First Prediction, Step by Step
Six steps from raw project data to ranked targets.
The first cycle (~six weeks) establishes the data foundation and modeling framework. Subsequent cycles reuse this foundation and complete in roughly two weeks each.
NDA & Kickoff
Confidentiality, project scope, commodity, boundaries, and deliverables confirmed.
Data Review
Drilling, assays, geophysics, GIS, interpretations, and gaps catalogued.
Data Transformation
Cleaning, alignment, and spatially informed feature construction.
AI Model Training
Hybrid data-driven and geology-aware models; 2D and 3D predictive outputs.
Interpretation & Ranking
Targets scored by geological rationale, data support, and confidence.
3D Model & Report
Delivery package: 3D model, ranked targets, technical report, exportable layers.
Full six-step foundation build. Most intensive phase of the engagement.
Steps 02–06 are repeated lightly: new data is folded in, models retrained, rankings revised.
Models reason around sparse, uneven, or biased data rather than overfitting dense areas.
What The Client Receives
Decision-ready outputs. Compounding value.
Core Deliverables
- Updated 3D predictive model
- Ranked exploration targets with confidence commentary
- Probability surfaces and quantitative prediction outputs
- Geological interpretation of model results
- Technical report and investor-ready summary
- Exportable layers for Leapfrog, Surpac, Vulcan, GIS, CAD
- Recommended field follow-up and drilling direction
Five Value Pillars
- Turning messy data into usable intelligence across drill logs, assays, maps, geophysics, and historical files.
- Reducing spatial and sampling bias from areas with dense drilling history.
- Capturing geological structure and continuity across proximity, orientation, alteration, and structural controls.
- Producing decision-ready targets, not just models: ranked, rationalized, uncertainty-quantified.
- Improving over time as new drilling and field evidence arrive each cycle.
The engagement is a six-month exploration intelligence program rather than a one-off report. Clients move from static interpretation to continuously improving guidance, and they always know where to spend their next drilling dollar.