Terranox AI
Product & Competitive Intelligence
AI-powered uranium exploration using physics-informed models on 70+ years of geology data.
Company Overview
Builds the first vertically integrated, AI-powered uranium exploration company that uses physics-informed machine learning models trained on 70+ years of legacy geological data to dramatically improve uranium discovery rates across North America.
Competitive Advantage & Moat
Product Roadmap & Public Announcements
Terranox has publicly stated they are building physics-informed AI models for uranium prospectivity mapping, digitizing decades of analog exploration records, and launching in-house exploration programs in North America. They've described a "compounding learning flywheel" where every drill result improves model accuracy, and a sequential decision intelligence engine that optimizes exploration spend. Their YC launch emphasized vertical integration, owning both the AI platform and the exploration projects it guides.
Signals & Private Analysis
The CTO's 8 years at RBC Borealis AI (rising to Head of AI/ML Systems leading 30+ people, shipping systems with $100M+ annual impact) and NASA JPL remote sensing background hints at satellite/hyperspectral imagery integration not yet publicly announced. Conference circuit appearances at mining-tech events suggest partnerships with Canadian provincial geological surveys for data access. The founders met in first-year physics and have been friends for 10+ years. Strong indicators of future SaaS licensing or JV model once field validation is complete, mirroring KoBold Metals' evolution but uranium-specific.
Product Roadmap Priorities
Physics-informed ML models analyze 70+ years of digitized geological, geophysical, and geochemical data to generate probabilistic uranium prospectivity maps that rank exploration targets by discovery likelihood.
It's like giving geologists a heat map that shows exactly where to dig for uranium instead of guessing across millions of acres.
It's like Waze for uranium—instead of wandering every back road hoping to avoid traffic, the AI already knows where the clear lanes are and routes you straight to the good stuff.
A compounding learning flywheel and sequential decision engine that optimizes every exploration action—where to drill, what to sample, which data to acquire—to maximize geological information gain per dollar spent.
It's like a chess engine for drilling—each move is calculated to learn the most about what's underground while spending the least money possible.
It's like playing 20 Questions with the Earth's crust, except the AI gets smarter with every answer and figures out where the uranium is hiding in half the questions.
Computer vision and NLP pipelines automatically digitize, extract, and structure 70+ years of analog geological records—hand-drawn maps, scanned PDFs, handwritten drill logs—into ML-ready geospatial datasets.
It's like teaching a computer to read your grandpa's handwritten geology notebooks and turn them into a searchable, map-ready database overnight.
It's like hiring a thousand interns who can read 1950s geologist handwriting, except these interns never sleep, never misfile anything, and automatically pin everything to a map.
Company Overview
Key Team Members
- Jade Checlair, Co-Founder & CEO
- Leeav Lipton, Co-Founder & CTO
Founder Force Multiplier
Terranox uniquely combines Jade Checlair's PhD in Geophysics from UChicago (published planetary discovery methods adopted by NASA flagship missions, 8 papers with 270 citations, ex-NASA Ames, 3.5 years at BCG leading nuclear and mining strategy) with Leeav Lipton's 8 years building AI/ML systems at RBC Borealis AI (Head of AI/ML Systems, 30+ person team, $100M+ annual impact) and ex-NASA JPL remote sensing expertise, giving them the rare ability to build physics-constrained ML models that actually respect geological reality.
Funding History
- 2025 | Jade Checlair and Leeav Lipton co-found Terranox AI.
- 2026 | Accepted into Y Combinator Winter 2026 batch.
- 2026 | Launching in-house exploration programs in North America.
Competitors
- AI-for-Mining Platforms: KoBold Metals (broader critical minerals, $500M+ raised), Earth AI (multi-mineral, Australia-focused).
- AI Geoscience Tools: Minerva Intelligence, Goldspot Discoveries (multi-commodity AI exploration).
- Traditional Uranium Explorers: Cameco, NexGen Energy, Denison Mines.
