Last week, we took the week off due to some operational commitments.

This week, were back with a new company profile: Mariana Minerals a company attempting to rebuild U.S. critical-minerals production under severe time, labor, and political constraints.

China controls the majority of global refining capacity for lithium, rare earths, and battery materials leaving U.S. defense modernization dependent on adversary-influenced supply chains.

The open question is whether coordination advantage is sufficient to overcome the structural realities of mining or whether software merely smooths the edges of an industry still governed by long cycle times.

Mariana Minerals is building a vertically integrated minerals company that owns and operates projects end-to-end, using proprietary software to coordinate permitting, construction, extraction, and refining.

The company’s wager is that better coordination, faster feedback loops, and tighter control over execution variance can meaningfully compress timelines and unlock marginal resources that traditional mining models leave stranded.

Founded in 2024 and headquartered in San Francisco with major engineering operations in Houston, Mariana emerged from the intersection of battery supply-chain experience and large-scale machine-learning systems.

CEO Turner Caldwell spent nearly a decade at Tesla working across factory design, battery materials, and upstream mining to develop techno-economic models for vertically integrated supply chains.  Mariana’s driving thesis is that mining and refining remain largely pre-digital industries, where decisions are still governed by spreadsheets, institutional knowledge, and static assumptions.

After securing an $85M Series A in July 2025, Mariana broke ground on its first commercial facility in October 2025. The site targets initial production by mid-2027, with planned capacity of approximately 3,000 metric tons per year of lithium salts.

MarianaOS functions as the company’s internal operating layer, integrating data and decision-making across the full project lifecycle. It is modular, composed of three primary systems aligned to distinct bottlenecks in minerals development:

Produced-water lithium extraction facility developed with Select Water Solutions, leveraging existing oil-and-gas wastewater infrastructure.

Mariana acquired the Lisbon Valley Mining Company in July 2025. The asset is a permitted, operating copper mine with FAST-41 transparency status.

Critical minerals have moved from a niche industrial concern to a headline national security issue. U.S. investments in MP Materials, sanctions exposure across defense supply chains, and $12B towards a strategic minerals stockpile all reflect a recognition of the costs and consequences of globalization.

At the same time, the United States is prioritizing everything at once: onshore manufacturing, rebuild shipyards, expand weapons production, dominate AI, field autonomy at scale, and harden cyber infrastructure.

On top of it, ever sector talks about the industrial workforce is aging and skilled labor is major constraint. But… when everything is a priority, nothing is.

If we actually want change, serious progress is going to require triage.

Under a time-constrained lens (2027 is 11 months away), some priorities sort themselves naturally. AI, cyber, and software-defined capabilities scale quickly. Drones can be manufactured and iterated on shorter timelines. These areas warrant sustained full pressure.

Heavy industrial rebuilds likely do not. Onshoring manufacturing, expanding shipyards, and opening new mines routinely take five to ten years before meaningful throughput is achieved even under favorable conditions.

Large, greenfield mining projects and shipyards will be strategically necessary over the long run, but they are poorly matched to near-term security demands.

As a result, a wholesale reversal of decades of globalized mineral production is unlikely to deliver decisive results on the timelines that matter most.

So if new mines cannot be stood up quickly, the remaining lever is to extract more value from what already exists: smaller deposits, waste streams, brownfield assets, and operations that were previously uneconomic due to coordination and recovery inefficiencies. Historically, these “long-tail” resources were ignored because scale economics dominated.

This is where Mariana Minerals fits.

Software does not change ore grade. It does not eliminate permitting timelines. It does not remove the need for physical capital. But Mariana’s thesis is strongest precisely where inefficiency is the binding constraint.

Under a triaged, near-term timeline, incremental gains may be the only gains available. Mariana does not need to “solve mining” to matter. It needs to make marginal projects viable faster and more predictably than traditional models allow.

Mariana Minerals is a serious attempt to modernize one of the most capital-intensive and coordination-constrained parts of the defense industrial base. Its software-first, vertically integrated model is directionally right and aligned with U.S. national security priorities, particularly where time, labor, and political constraints limit traditional industrial rebuilds.

Owning and operating physical assets concentrates capital, permitting, and organizational challenges that software alone cannot remove. The model only works if Mariana can materially compress timelines and reduce execution variance enough to make marginal assets viable faster than incumbent approaches. 

If it does, it offers a credible path to a new class of prime-scale minerals operator. If not, it will reaffirm the structural limits of software-driven transformation in heavy industry.