Mining Intelligence

The intelligence foundation AI-native mining requires.

We build the platform and applications that connect orebody intelligence to operating decisions — capturing value that legacy mine systems cannot access.

Learn More

The Problem

Most mining data never reaches a decision.

Modern mines generate substantial sensor data across geology, equipment, and processing. That data is rarely unified in time to influence decisions, and rarely connected to the systems responsible for production outcomes.

The data used to plan a mine is modeled from sparse drill holes for long-range planning — not for the specific ore entering the plant. When operations begin, that planning model is rebuilt through short-term grade control. This creates the Data Chasm — where the ore being processed is characterized by a different dataset than the one used to plan its extraction.

The data chain is broken.

The consequences are measurable and consistent. Across large-scale operations, plan-to-actual variance of more than 20% is the norm — not the exception. One in five tonnes of economically recoverable metal is never extracted at full value. Industry-wide, $50–100 billion in annual economic value is at stake.

>20%

Plan-to-actual variance across large-scale operations

1-in-5

Tonnes of recoverable metal never extracted at full value

$50–100B

Annual economic value at stake industry-wide

What We Do

Alloy builds the foundation that AI-native mining requires.

The structural break between geological signal and plant performance — the Data Chasm — is the root cause of systemic variance in mining. AI systems built on top of this broken data chain cannot fix it; they inherit its limitations. Alloy builds the layer beneath.

AI trained on low-fidelity data delivers low-fidelity insight.

The Alloy Intelligence Platform assigns a continuously updated, persistent identity to ore — from drill core through extraction and processing — associating raw geological signal directly with downstream outcomes. By ingesting and preserving high-fidelity sensor data without the compression legacy systems require, Alloy ensures that orebody intelligence drives every operational decision.

On top of that foundation, we deploy applications that turn intelligence into action: ore tracking, material routing optimization, plant stability, and autonomous decision execution within established operating parameters. As mines transition toward multi-agent AI environments, Alloy maintains common context across all of them — ensuring every system operates from a single, consistent foundation.

The Team

Built by the team that identified the problem.

The Alloy founding team has independently built and proven the core capabilities the platform now integrates — across AI, spectral geology, sensor science, data science, and enterprise infrastructure.

JT Clark

CEO & Founder

Nearly two decades building AI systems in mining — first at BCG, where he developed a real-time AI mill optimizer that delivered 7–15% production improvement at a major copper operation, and then at Veracio, where he worked at the data layer itself. Alloy is the integration of those two experiences. Two granted AI patents, three pending. MBA, Wharton.

James Shreeve

Chief Client Officer

Recognized as a global authority on the application of sensory technologies to orebody understanding. At Geotek, built a multi-million dollar core scanning enterprise and developed BoxScan, now deployed globally. As EVP Operations at Veracio, led field delivery across major mining jurisdictions. MRes, University of Southampton.

The team collectively holds six granted patents and three pending across ore characterization, sensing, and applied machine learning.