The most expensive part of geothermal power is the first mile. Drilling a well costs millions, and if you hit cold rock, you get nothing but a dry hole and a write-down. For decades, this exploration risk has kept geothermal a niche player, confined to obvious hot springs and volcanic zones. Zanskar’s bet is that you can shrink that risk to almost nothing with a sufficiently smart map.
Founded in 2021, the Salt Lake City company started as a software play, using AI to comb through public geological data and find overlooked heat sources. The idea was neat, but the unit economics of selling software to other drillers were uncertain. So Zanskar did the logical, capital-intensive thing: it bought a power plant. In 2024, it acquired the Lightning Dock geothermal facility in New Mexico, the state’s only utility-scale plant [thinkgeoenergy.com, 2026]. Then it used its own models to pick a spot for a new well, which turned into what the company calls a “world-beating gusher” [zanskar.com]. The software wasn’t just a product anymore; it was the edge for an operator.
From SaaS to Steam
Zanskar’s pivot is a classic climate tech story of vertical integration. The initial wedge was pure data science: applying machine learning to massive geophysical datasets to identify “blind” geothermal systems that traditional surveys missed [Caplight]. The company positioned itself as a service for utilities and independent developers, helping them minimize exploration risk [Caplight]. But the path to scaling that model ran into the same wall every infrastructure software company hits: your customers move slowly, and their success is not your equity.
The answer, backed by a $115 million Series C round led by Spring Lane Capital in May 2025, was to become the developer [Yahoo Finance, May 2025]. Zanskar is now using its capital to rapidly build multiple geothermal power plants across the western United States [thinkgeoenergy.com]. The software is the proprietary advantage, but the asset is the power plant. This shifts the business from a technology licensing fee to selling firm, carbon-free electricity under long-term contracts,a fundamentally different, and potentially much larger, revenue model.
The Traction: A Plant and a PPA
The proof points for this capital-heavy shift are now on the board. Beyond the Lightning Dock acquisition, Zanskar has signed a 20-year power purchase agreement with utility NV Energy for 100 megawatts of geothermal baseload power in Nevada [renewableenergymagazine.com, 2026]. For context, that’s enough electricity to power roughly 80,000 homes, around the clock. The company claims to have used its AI tools to uncover the specific blind geothermal system that will feed this project in the western Nevada desert [Canary Media, 2025].
| Milestone | Detail | Source |
|---|---|---|
| Lightning Dock Acquisition | Acquired New Mexico's only utility-scale geothermal plant in 2024. | [thinkgeoenergy.com, 2026] |
| NV Energy PPA | Signed 20-year deal for 100 MW of baseload power in Nevada. | [renewableenergymagazine.com, 2026] |
| Series C Funding | Raised $115 million in May 2025 led by Spring Lane Capital. | [Yahoo Finance, May 2025] |
| Resource Discovery | AI identified a blind geothermal system in Nevada with >100 MW potential. | [Canary Media, 2025] |
Co-founders Carl Hoiland (CEO) and Joel Edwards (CTO) lead the company, which has attracted a who’s who of climate-focused investors including Obvious Ventures, Union Square Ventures, Lowercarbon Capital, and Munich Re Ventures [Latitude Media]. The investor list suggests confidence not just in the AI, but in the team’s ability to navigate the complex, multi-year process of permitting, drilling, and connecting a power plant to the grid.
Where the Wheels Could Come Off
Building power plants is a different game from selling software, with different risks. The capital required is orders of magnitude larger, and development timelines are measured in years, not quarters. Zanskar is now competing directly with established geothermal developers and other clean energy projects for grid interconnection slots, steel, drilling rigs, and engineering talent.
- Execution Risk. The step from a successful exploration well to a fully permitted, constructed, and operational power plant is a marathon with countless potential delays. A single regulatory hiccup or construction overrun could burn through capital.
- Commodity Competition. Geothermal must compete on cost not just with other renewables, but with cheap natural gas and increasingly affordable grid-scale batteries. Zanskar’ mission is to make it the most affordable form of firm clean power, but that’s a promise that will be tested at the meter [zanskar.com].
- Technology Scaling. The AI’s success in identifying one or two promising sites must be replicated across a portfolio of projects to achieve the gigawatt-scale pipeline the company claims to own [ESG Tech Report, 2026].
The company’s most plausible answer to these risks is that its AI-driven exploration drastically improves the odds of success from the start, lowering the average cost per megawatt over a portfolio of projects. If the software can consistently find hotter, more productive resources with fewer dry holes, it changes the underlying economics of the whole category.
The Next Twelve Months
The immediate focus is turning signed paper into steam. The NV Energy PPA is a landmark, but it’s a contract to deliver power, not delivered power. Zanskar will need to progress its Nevada project through detailed resource confirmation, permitting, and construction. Watch for announcements of a final investment decision or the start of drilling. Similarly, further expansion of the Lightning Dock facility in New Mexico would demonstrate an ability to repeatedly develop resources.
Another signal will be the company’s next financing move. The $115 million Series C is substantial, but building multiple plants will require even more capital, likely in the form of project finance debt. Zanskar’s ability to attract non-dilutive capital from banks or infrastructure funds will be a key test of its model’s bankability.
On the back of an envelope, the bet looks like this: if Zanskar’s AI can improve the success rate of geothermal exploration from, say, 50% to 80%, it cuts the exploration risk premium embedded in every project’s cost of capital. For a 100 MW plant costing $500 million to build, that could mean tens of millions in saved financing costs before a shovel even hits the ground. That’s the margin they’re chasing.
To win, Zanskar doesn’t just have to be better than other AI geologists. It has to be better than Fervo Energy, the current darling of next-generation geothermal, which has also landed major utility contracts and is pushing down costs through advanced drilling techniques [Competitors]. Zanskar’s angle is the map; if its map is truly superior, it could own the best land. If not, it’s just another developer in a very hard, very expensive business.
Sources
- [Yahoo Finance, May 2025] Zanskar raises $115M Series C to accelerate geothermal development | https://finance.yahoo.com/news/zanskar-raises-115m-series-c-130000291.html
- [thinkgeoenergy.com, 2026] Zanskar acquires Lightning Dock Geothermal power plant | https://www.thinkgeoenergy.com/zanskar-acquires-lightning-dock-geothermal-power-plant/
- [renewableenergymagazine.com, 2026] Zanskar signs 100 MW PPA with NV Energy | https://www.renewableenergymagazine.com/geothermal/zanskar-signs-100-mw-ppa-with-nv
- [Latitude Media] Armed with $115 million, geothermal startup Zanskar gets ready to build | https://www.latitudemedia.com/news/armed-with-115-million-geothermal-startup-zanskar-gets-ready-to-build/
- [Caplight] Zanskar - Company Profile | https://www.caplight.com/company/zanskar
- [zanskar.com] Progress at Lightning Dock: A Big Update | https://www.zanskar.com/blog/betting-on-an-underperforming-geothermal-resource-zanskars-new-step-out-production-well-at-lightning-dock-is-a-world-beating-gusher
- [Canary Media, 2025] How AI is helping unearth hidden geothermal resources | https://www.canarymedia.com/articles/geothermal/how-ai-is-helping-unearth-hidden-geothermal-resources
- [ESG Tech Report, 2026] Geothermal Startup Zanskar Raises $115 Million | https://www.esgtoday.com/geothermal-startup-zanskar-raises-115-million-to-find-and-develop-carbon-free-energy-using-ai/