
Mariana Minerals builds and operates mines and refineries using an integrated software platform.
Mariana Minerals offers MarianaOS, an AI-native mineral intelligence and operations platform, alongside two flagship projects: Copper One and Lithium One. MarianaOS integrates geoscience data, mine planning, capital project execution, and autonomous plant control into a single software stack designed to accelerate the path from discovery to production.
Copper One is an operating copper mine and refinery in Utah that Mariana acquired in late 2025, where it is deploying autonomous mining and refining technologies to scale output. Lithium One is a GWh-scale facility designed to produce battery-grade lithium salts from oil and gas produced water, with first commercial production targeted for the first half of 2027.
Demand for copper, lithium, nickel, cobalt, and other critical minerals is being driven by electrification, data-center and AI infrastructure expansion, renewable-energy deployment, and defense supply-chain reshoring. These structural trends are widening the gap between projected demand and the Western world's ability to permit, build, and operate new mines and refineries.
Mariana Minerals is positioned to address this supply shortfall through a software-first, vertically integrated model that uses its MarianaOS platform and autonomous operations to accelerate project development and reduce production costs across copper and lithium assets in the United States.
Mariana's core advantage is its software-first operating model, anchored by MarianaOS, which unifies exploration, capital project delivery, mining, and refining on one AI-driven platform. This integration lets the company compress timelines, reduce labor intensity, and optimize operations in ways legacy miners cannot match.
The company also combines in-house engineering, construction, and operating expertise with a vertically integrated project portfolio, giving it direct control over execution rather than relying on third-party contractors. Early deployment at Copper One and Lithium One provides real-world validation and a data feedback loop that should improve model performance and operational outcomes faster than competitors limited to software-only offerings.
Mariana faces the capital intensity and long lead times inherent to mining and refining, where even software acceleration cannot fully eliminate permitting, construction, and ramp-up risk. The company is also entering markets dominated by entrenched incumbents and state-backed Chinese producers with lower costs and decades of operational experience.
Its autonomy-first approach is still early: deploying reinforcement learning and autonomous systems at operating mine and refinery sites involves technical, regulatory, and workforce-transition risks. A tight Western labor pool for mining and metallurgical expertise, plus volatile commodity prices, could strain project economics and execution speed even if the technology performs as planned.