Zanskar Geothermal: The AI Prospector Hunting Hidden Geothermal Systems — and Whether You Can Buy Its Stock

Almost every geothermal plant running in the American West sits where the Earth tipped its hand — a hot spring, a steam vent, ground that stays bare in January snow. That's how prospectors found the resources a century ago, and the industry has been drilling next to those giveaways ever since.

The catch: most of the heat doesn't advertise. Geoscientists call the hidden reservoirs "blind" systems — hot, permeable rock with no surface expression at all. Nobody knows how many exist, because by definition you can't see them from above. Zanskar, a venture-backed company out of Utah, has built its whole business around a bet: that machine learning can find them, and that the company can then drill and develop what its models flag.

If you landed here asking whether you can buy Zanskar stock, that answer comes first, and it's short. Then the company, the technology, the money, and the projects.

Is Zanskar Publicly Traded?

No. Zanskar is a privately held, venture-backed company. There is no ticker symbol, no exchange listing, and no announced IPO. You cannot buy Zanskar shares through a brokerage account, and any site showing a "Zanskar stock price" is confusing it with something else.

The ownership picture is standard late-stage venture capital: its disclosed backers include Union Square Ventures, Lowercarbon Capital, Obvious Ventures, and — leading the most recent round — Spring Lane Capital. Private company, no public shares.

If you want geothermal exposure you can actually buy, there are two useful next stops. Fervo Energy went public in May 2026 and trades on the Nasdaq — a different company on a different technical approach. And our geothermal stocks and investing guide walks through the listed pure plays, the adjacent picks, and the honest limits of the whole category.

What Zanskar Actually Is

The full legal name is Zanskar Geothermal & Minerals, Inc. Its co-founders, CEO Carl Hoiland and CTO Joel Edwards, both hold geoscience PhDs, per the company's own story page. Startup databases date the founding to 2019, though the company doesn't publish an incorporation date. Headquarters: Zanskar says "along the Wasatch Front" in Utah; its federal SBIR profile lists a Provo address.

What the company does is broader than the AI label suggests. Zanskar describes a four-stage model: discover geothermal prospects, validate them with fieldwork and exploration drilling, develop power projects, and ultimately own and operate the plants. It calls itself the "first AI-native geothermal energy company" — its phrase, not ours — and, by its own account, it operates a power plant, Lightning Dock in New Mexico (details below).

That last part matters for reading the company correctly. Plenty of startups sell exploration software or survey data and stop there. Zanskar's stated intent is to develop, build, own, and operate — which makes it a power-plant developer with an unusual front end, not a software vendor.

By its own description, everything it targets is conventional hydrothermal: naturally hot water circulating through naturally permeable rock, the resource type behind most U.S. geothermal electricity generated today. Keep that in mind — it's the key to what the company is and isn't claiming.

The Technology: Finding Hidden Heat

Here's why this is a genuinely hard problem, and why it attracted a nine-figure funding round.

A geothermal power plant needs three things underground at the same spot: heat, fluid, and permeability. Heat alone is everywhere if you drill deep enough — temperature generally increases with depth — but "deep enough" is usually too expensive. A commercial hydrothermal resource is the lucky intersection where hot fluid moves through fractured rock at a depth you can afford to reach.

For a century, the industry found those intersections by walking toward the steam. Systems that leak to the surface got found; systems that don't, mostly didn't. Blind systems have historically turned up by accident, and testing a hunch means drilling wells that cost millions each. Exploration cost and dry-hole risk are among the biggest barriers that have kept conventional geothermal small.

Zanskar's answer, as the company describes it: aggregate every kind of subsurface and remote-sensing data it can get — geological, geophysical, thermal, satellite — feed it to machine-learning models that predict where heat and permeability coincide, verify the strongest predictions with field measurements, and only then spend drilling money on the highest-probability targets.

That's the company describing its own product. Treat it as a claim, not a finding. But two outside data points give it more texture than the usual AI pitch.

First, a discovery with independent press coverage. In December 2025, Axios reported on Zanskar's Big Blind find in western Nevada — a geothermal system with no visible surface signs at all. Per that reporting, the company's models flagged an anomaly associated with high heat flow, and its prediction engine narrowed the drilling targets; Zanskar executives told Axios the goal is fewer failed wells and more precise targeting.

Second, a federal paper trail. The Department of Energy's SBIR database — a government record, not a press release — shows $449,849 in Phase I awards, $1,614,946 in Phase II, and $2,064,795 total for Zanskar. The funded work is concrete: using machine learning to predict naturally fractured reservoirs before wells are drilled, and deep-learning denoising of distributed acoustic sensing (DAS) data to sharpen seismic signals for reservoir characterization. That's grant money, not venture funding, and it's small next to the equity rounds. It proves the awards were made for the work described — twice — not that DOE endorsed the commercial thesis.

What Zanskar Is Not

Worth pinning down, because "AI plus geothermal plus Nevada" pattern-matches onto several companies doing very different things.

Zanskar's public positioning centers on natural reservoirs, not engineered ones. Enhanced geothermal systems manufacture their own reservoir — drilling into hot rock that lacks natural permeability and fracturing it so fluid can flow. That's Fervo Energy's approach, and it's why Fervo drills where the resource is known but locked. Zanskar's published work runs the other way — discovery, not reservoir creation. Its bet is that nature already built plenty of reservoirs; we just haven't found them. Same destination — firm, carbon-free power — reached through opposite technical claims.

Its public materials emphasize discovery, not a new power-conversion cycle. Once a resource is found and drilled, the plant on top is conventional. The AI's job, in Zanskar's framing, is finding the resource — not extracting anything.

And it's one company, not a category. Whether Zanskar succeeds or fails won't settle whether AI-assisted exploration works in general, and nothing here describes the geothermal industry at large. This is a profile of a single, privately held developer with a specific claim.

The Funding Record

Here's the round-by-round record, with the sourcing quality of each figure stated plainly — private-company funding is announced, not audited, so the attribution matters.

The Series C is the headline: $115 million, announced January 21, 2026, led by Spring Lane Capital — a round Zanskar says brings its total equity funding to date to $180 million. Every earlier lead came back for it: Obvious Ventures from the Series B, Union Square Ventures and Lowercarbon Capital from the Series A, plus Munich Re Ventures and, per the company, more than a dozen other energy and technology investors. Repeat investors writing bigger checks is generally a healthier signal than a new lead swooping in alone — though it's still the company's announcement, not an audited figure.

Notice the arithmetic, though. The three named rounds add up to $157 million, not $180 million. Zanskar's public round announcements don't fully account for its own stated total — common enough for startups, which raise unannounced extensions and convert notes between lettered rounds — but it means the round-by-round record and the headline number don't reconcile on paper. Worth knowing before you quote either one.

Keep the DOE money in its own column. The $2,064,795 in SBIR awards is federal grant and contract support for specific research projects. It is not equity, it doesn't belong in a "total raised" figure, and adding it to the venture rounds would overstate both numbers' meaning.

Projects and the NV Energy PPA

Lightning Dock — New Mexico

Zanskar says it purchased the existing Lightning Dock geothermal plant in southern New Mexico in May 2024, found the underperforming field's problem, drilled into a deeper and hotter zone, and restored the plant to full capacity in under a year. The company calls it New Mexico's only utility-scale geothermal plant and says one of its wells there is the most productive pumped geothermal well in the United States. All of those are Zanskar's claims. No independently verified capacity figure for the plant is public, and the company doesn't publish one. What the acquisition does show: this is a company that operates real hardware, not just models.

Big Blind — western Nevada

The discovery that made Zanskar's case in public. Announced in December 2025 and covered by Axios, Big Blind is a hydrothermal system in remote western Nevada with no surface signs of geothermal activity — found, per the company, by its prediction models. Zanskar puts the resource around 250°F at roughly 2,700 feet — the company's figures, not an independent measurement — shallow by geothermal drilling standards. It is a discovery, not a power plant: permitting comes next, and Zanskar has said it hopes the site can deliver power later this decade.

Pumpernickel — northern Nevada

A named discovery in Zanskar's Nevada portfolio, cited by the company among its recent finds. Public details are thin — no published capacity, timeline, or project-specific offtake that we could verify.

The 100 MW NV Energy PPA

In May 2026, Zanskar announced a 20-year power purchase agreement with NV Energy for 100 MW of firm geothermal power. Two qualifiers belong right next to that number. It was announced as conditional — deliveries depend on the Public Utilities Commission of Nevada approving NV Energy's 2026 Integrated Resource Plan. And the announcement did not specify which of Zanskar's projects would supply the 100 MW. A PPA with a major utility is a real commercial milestone — and not yet a revenue stream.

Frequently Asked Questions

Is Zanskar publicly traded? What's the stock symbol? No, and there isn't one. Zanskar is privately held and venture-funded, with no ticker, no exchange listing, and no announced IPO. Ordinary investors cannot buy shares.

Who founded Zanskar, and where is it based? Co-founders Carl Hoiland (CEO) and Joel Edwards (CTO), both geoscience PhDs, per the company's story page. It's based in Utah — the company says the Wasatch Front; its federal SBIR profile lists a Provo address.

Is Zanskar an EGS company like Fervo? Not in its public positioning. Fervo engineers artificial reservoirs in hot rock; Zanskar's published work is about finding natural reservoirs hidden from the surface, then developing them conventionally. Our EGS explainer covers the difference in depth.

How much money has Zanskar raised? Zanskar says its January 2026 Series C brought total equity funding to date to $180 million. The announced rounds: a $12 million Series A co-led by Union Square Ventures and Lowercarbon Capital (August 2022), a $30 million Series B led by Obvious Ventures (May 2024), and a $115 million Series C led by Spring Lane Capital (January 2026). Those total $157 million — the remainder of the stated $180 million isn't publicly itemized. Separately, federal records show $2,064,795 in DOE SBIR grant awards.

What is Big Blind? A hydrothermal system in western Nevada with no surface signs, which Zanskar announced in December 2025 as a discovery made by its AI prediction models. It's in early development — permits first, power later this decade if it works out.

The Bottom Line

Zanskar's pitch is that geothermal's real bottleneck is discovery — that the resource base is much bigger than the map of hot springs suggests, and that models trained on geological, geophysical, and satellite data can find the hidden part cheaper than trial-and-error drilling ever could. The record behind that pitch: repeat DOE research funding, an operating plant it bought and says it fixed, one blind discovery with credible independent coverage, escalating venture rounds that the company says total $180 million in equity, and a conditional 100 MW offtake with a major utility.

What's not yet on the record is the proof that matters most — commercial power flowing from a resource nobody knew existed until a model said to drill there. That's the claim the whole company is built on, and as of this writing it's still ahead of them. Worth watching either way: if the discovery math works, the map of American geothermal gets a lot bigger than the steam ever let on.