AI Prospectivity Pipeline
Clay Foundation Model → Predictive Kernels → Mineral Discovery
6-Stage Pipeline
Ingest
Multi-sensor satellite data ingestion (Sentinel-2, Landsat, EMIT)
Embed
Clay Foundation Model generates 1024-dim geological embeddings
Analyze
Spectral index computation and mineral signature detection
Predict
ML classifiers generate mineral prospectivity heat maps
Protect
DMIR registration and Royale Protocol sovereignty layer
Query
ATLAS Agent natural language interface for intelligence queries
Clay Foundation Model
Open-source Earth observation foundation model by Radiant Earth. Generates geological embeddings from multi-sensor satellite data.
Spectral Indices
Iron Oxide Ratio
Band 4 / Band 2Detects iron-rich surface expressions indicative of mineralization
Clay / Hydroxyl Index
Band 11 / Band 12Identifies alteration zones associated with hydrothermal deposits
Ferrous Minerals
Band 6 / Band 8AMaps mafic host rocks for PGM, nickel, and chromite deposits
Carbonate Index
Band 13 / Band 14Identifies carbonatite complexes associated with REE and niobium