Terranox AI
AI-powered uranium exploration
About
Terranox is the first vertically integrated AI-powered uranium exploration company. We find high-grade uranium deposits in North America using AI trained on 70+ years of exploration outcomes. Our mission is to find the fuel we need to power the next century with nuclear energy.
Founders
Founder
Co-founder, CEO of Terranox AI, the first vertically integrated AI-powered uranium discovery company. PhD in Geophysics from UChicago, ex-NASA Ames and ex-BCG mining & energy.
AI Research Report
Problem & Solution
Problem & Solution — Terranox AI
Terranox AI addresses a critical bottleneck in the global transition to clean energy: a looming uranium supply shortfall. As nuclear energy expansion gains momentum through policy shifts and corporate commitments (such as hyperscalers seeking carbon-free power for AI datacenters), the demand for uranium is projected to grow multi-fold by mid-century. However, the exploration industry is currently ill-equipped to meet this demand. Traditional exploration workflows are slow, heavily reliant on human intuition, and suffer from hit rates well below 1%. Furthermore, decades of valuable geoscience data remain trapped in fragmented, non-digital formats like PDFs and physical maps.
The significance of this problem is compounded by the long lead times required to bring a new mine from discovery to production, which typically ranges from 10 to 15 years. This delay creates a strategic risk for national energy security and utilities. Terranox identifies this inefficiency as an opportunity to apply modern machine learning to accelerate the discovery of economic deposits in stable jurisdictions, specifically focusing on North America.
Terranox’s solution is to become the first vertically integrated, AI-powered uranium discovery company. Their technical stack consists of three core pillars. First, they utilize multi-modal geoscience intelligence to extract and process 70+ years of fragmented historical data. Second, they employ prospectivity mapping using AI models specifically trained on historical uranium exploration outcomes. Third, they use sequential decision intelligence to optimize field actions, ensuring that every dollar spent on drilling maximizes information gain through an active learning loop.
The value proposition for the mining industry is a significant improvement in discovery efficiency, leading to higher hit rates and reduced capital waste. For energy stakeholders, it offers a faster path to domestic uranium supply. By combining predictive modeling with their own field programs, Terranox creates a 'compounding learning flywheel' where every drill result—whether a hit or a miss—improves the accuracy of their models for future projects.
Market & Competitors
Market, Trends & Competitive Landscape — Terranox AI
The uranium market is entering a period of renewed growth, driven by the global push for decarbonization and the massive energy requirements of AI datacenters. Organizations such as the IAEA and the World Nuclear Association forecast material demand growth over the coming decades. This has led to a multi-year upcycle characterized by rising uranium prices and a significant increase in capital spending by mining companies. S&P Global notes that aggregate uranium revenue for major producers is estimated to triple between 2023 and 2033, creating a fertile environment for new exploration technologies.
The competitive landscape for Terranox AI is multi-faceted. It includes incumbent 'majors' such as Cameco, Kazatomprom, and Orano, who dominate current supply. It also includes well-funded developers like NexGen Energy and Denison Mines. While these incumbents have vast resources, they often rely on traditional exploration methods. Terranox also competes with a large number of junior exploration companies that focus on prospecting in North America.
In the specialized niche of AI-driven exploration, Terranox faces competition from other vertical AI startups. A notable example is Earth AI, which also uses a vertically integrated model (AI plus hardware/drilling) for mineral discovery. Additionally, broader data-first exploration firms like KoBold Metals represent a potential threat if they choose to expand their focus more aggressively into the uranium sector.
Terranox’s competitive advantage lies in its uranium-specific AI models and the deep domain expertise of its founders in both geophysics and nuclear strategy. Their vertically integrated approach allows them to capture more value than a pure software provider. However, they face significant risks, including the extreme capital intensity of drilling programs and the long regulatory timelines associated with mining. Furthermore, the market is highly concentrated, meaning that successful discovery must eventually navigate the deal dynamics of a few very large producers and utilities.
Total Addressable Market
Quantitative TAM and Market Metrics — Terranox AI
Terranox AI operates at the intersection of uranium discovery and artificial intelligence. To understand its market potential, two Total Addressable Market (TAM) definitions are relevant: the global uranium fuel commodity market and the specific addressable expenditure on exploration and upstream capital projects.
According to the World Nuclear Association (WNA), global reactor requirements for uranium in 2025 are estimated at approximately 68,920 tonnes of uranium (tU). When converted to pounds (roughly 152 million lbs) and evaluated against industry price forecasts, the implied annual market value is significant. Using a price range of $75/lb to $95/lb (based on S&P Global and Visible Alpha projections of rising realized prices), the 2025 uranium fuel market is valued between $11.4 billion and $14.4 billion. This represents the total value of the commodity that Terranox's exploration technology aims to locate.
A more direct measure of Terranox's Serviceable Addressable Market (SAM) is the upstream capital expenditure (capex) dedicated to exploration and development. S&P Global and Visible Alpha report that aggregate uranium capex is expected to rise from $704 million in 2024 to $969 million in 2025, eventually peaking near $1.6 billion in 2027. Terranox targets the portion of this spend allocated to exploration, data services, and decision intelligence.
The methodology for these estimates relies on multiplying projected reactor demand by forecasted realized prices for the commodity TAM, while the exploration TAM uses reported industry capex as a proxy for available spending. While the commodity market is in the tens of billions, the immediate addressable market for AI-driven exploration services is in the low hundreds of millions to low billions annually.
It is important to note that realized prices can vary significantly between long-term contracts and spot markets. Additionally, exploration budgets are often project-dependent and can be volatile. However, the accelerating capex trend indicates a growing market for efficiency-driving technologies like those offered by Terranox.
Founder Analysis
Founders and Background — Terranox AI
Jade Checlair is the Co-founder and CEO of Terranox AI. She holds a PhD in Geophysics from the University of Chicago and has a distinguished background in both academia and industry. Her prior experience includes developing statistical methods for planetary discovery at NASA Ames, which were subsequently adopted by NASA flagship missions. Before founding Terranox, she spent approximately 3.5 years at Boston Consulting Group (BCG), where she led nuclear and mining strategy, bridging the gap between complex geosciences and commercial strategy.
Leeav Lipton serves as the Co-founder and CTO. He brings approximately eight years of experience building AI and machine learning systems, notably at Borealis AI. His academic background is in astrophysics, and he previously worked as a scientist at NASA's Jet Propulsion Laboratory (JPL), focusing on low-cost remote sensing technology. This combination of deep engineering experience in production ML and a background in remote sensing provides the technical foundation for the company's predictive models.
The founders share a long history of collaboration, having met during their first year of physics studies and worked together for over a decade. This long-term partnership, combined with their complementary skill sets—Jade’s geophysics and strategic mining expertise and Leeav’s AI engineering and remote sensing background—positions the team as domain-led technical leaders. Their credibility is further bolstered by the company's acceptance into the Y Combinator W26 cohort.
Currently, the company operates with a lean team of 2–10 employees. The leadership team maintains direct engagement with the industry, utilizing a dedicated outreach channel for founder-level inquiries. Their combined experience at NASA and in top-tier consulting and AI research labs provides a unique 'signal' of technical excellence and market understanding in the specialized field of uranium exploration.
Unlock Full AI Research Report
Enter your email to access the complete analysis.
We'll never spam you. Unsubscribe anytime.