The most expensive part of a new battery storage project isn't the steel or the cells. It's the time spent waiting for a permit, a process that can stretch for years and cost millions in lost revenue. Permeta.ai, a Portland-based startup, is betting that the best way to accelerate a $100 million solar farm isn't a bigger bulldozer, but a smarter piece of software that checks your paperwork before you even file it [8 Clockwise, September 2026].
Founded in 2025, the company is building what it calls a patent-pending neuro-symbolic AI platform designed to evaluate permit applications as they are being developed. The goal is to give project developers and regulators greater clarity and consistency, reducing the back-and-forth that bogs down energy and industrial infrastructure [8 Clockwise, September 2026]. It's a classic climate tech play: focus on the unit economics of time. Every month shaved off a permitting delay translates directly into earlier revenue for a wind farm or avoided capital costs for a carbon capture facility.
A wedge through carbon capture
Permeta.ai didn't start with a broad brush. Its initial product, launched under the name ClassVI.ai, was narrowly targeted at the complex permitting process for carbon capture and sequestration wells [Permeta.ai, Unknown]. This was a smart wedge. The regulatory framework for Class VI wells is notoriously dense, a perfect proving ground for an AI that needs to ingest and interpret structured rules. Having established a beachhead there, the company has since expanded its scope to include liquefied natural gas, utility-scale solar, and battery storage projects [Yahoo Finance/ACCESS Newswire, June 2026]. The throughline is clear: high-stakes, heavily regulated energy infrastructure where delays are measured in millions, not days.
The platform serves a two-sided market, aiming to streamline workflows for both project developers and the regulators reviewing their submissions [Permeta.ai, Unknown]. An early engagement reportedly made the Federal Energy Regulatory Commission pre-filing process up to 25% more efficient for a customer [Startuply.vc, Unknown]. The company has also announced a partnership with Coastal Bend LNG, providing generative AI tools to accelerate their FERC pre-filing and permit preparation [Caroline Soliman, LinkedIn, 2026].
The team from Nike and Big Oil
What makes this bet more than just another regulatory tech play is the founding team's blend of deep tech operational experience and energy domain expertise. It's a group that looks less like a typical seed-stage startup and more like a strategic spin-out.
| Role | Name | Key Background |
|---|---|---|
| Co-Founder | Eric Redmond | 30-year tech veteran, six-time founder, co-founded Nike's Tech Innovation Office and first AI team [Permeta.ai About Us, Unknown]. |
| Co-Founder & Chief Customer Officer | Heidi Eggert | Co-founded Nike's Tech Innovation Office, adviser to the World Economic Forum, founder of Planning Team 11 [Permeta.ai About Us, Unknown]. |
| Chief Product Officer | Anya Shvilpe | Energy and carbon-management specialist with leadership roles at Wood Mackenzie, Halliburton, and BP [Permeta.ai About Us, Unknown]. |
| Chief Financial Officer | Carl Stumpf | Experience at UI Solutions, Accenture, CME Group, and Nike Technology [Permeta.ai About Us, Unknown]. |
| VP of Engineering | Roger Ignazio | Former Nike director specializing in AI/ML, automation, and systems engineering [Permeta.ai About Us, Unknown]. |
The Nike connection is strong, with multiple leaders hailing from its tech innovation groups. This suggests a team accustomed to shipping complex software at scale. Shvilpe's background provides the crucial domain knowledge, ensuring the AI isn't just technically clever but practically useful for navigating agencies like the FERC or state environmental departments.
The outcome-based pricing gamble
In a market saturated with seat-based SaaS, Permeta.ai is attempting a different commercial model. Chief Financial Officer Carl Stumpf has said the company's pricing is based on outcomes, such as a specific permit type, rather than per-user seats [Carl Stumpf, LinkedIn, June 2026]. Customers can also add or remove users during active permitting periods, aligning cost directly with project phases [Carl Stumpf, LinkedIn, June 2026].
This is a bold and logical move for the space. A developer cares about getting a permit approved, not how many logins they purchased. If Permeta can tie its fee to a successful, accelerated outcome, it captures value in line with the immense value it provides. The risk, of course, is accurately pricing that outcome and managing the potential liability if a project still gets delayed for reasons outside the software's control.
Where the wheels could come off
For all its promise, Permeta.ai faces significant headwinds. The regulatory landscape is not just complex, it's fragmented across thousands of federal, state, and local jurisdictions. An AI trained on federal FERC guidelines might stumble on a county-level zoning ordinance.
- The customization trap. The "neuro-symbolic" approach implies combining learned patterns with hard-coded rules. Every new jurisdiction or project type could require expensive, manual rule-setting, threatening margins.
- The adoption friction. Regulators are historically slow to adopt new technology. Permeta's two-sided strategy is wise, but convincing cautious public agencies to rely on a startup's AI will be a long, relationship-heavy sales cycle.
- The invisible competitor. The most formidable incumbent isn't another software company. It's the entrenched ecosystem of law firms, environmental consultants, and manual processes that have dominated permitting for decades. Their product is risk mitigation, and they have deep, trusted relationships.
The company's answer appears to be a focused, use-case-by-use-case expansion, starting with the most standardized federal processes. The recent seed investment from the 8 Clockwise Seed Fund, which focuses on industrial and energy software, provides capital and credibility to grind through that long game [8 Clockwise, September 2026].
The next twelve months
Permeta.ai's immediate roadmap will be about proving repeatability. The Coastal Bend LNG partnership is a key reference customer. The next milestones will likely be announcing similar deals in the solar and battery storage verticals, and potentially a state-level regulatory pilot. Given its seed stage and small team size (estimated at 1-10 employees), the next 12 months are about moving from promising early engagements to a handful of solid, referenceable enterprise contracts [LinkedIn company profile, Unknown].
On the back of an envelope, the economic case is compelling. If a 500-megawatt battery storage project loses $50,000 in potential revenue every day it's delayed, shaving six months off a two-year permitting process saves over $9 million. If Permeta.ai can credibly claim even a fraction of that time savings, its outcome-based fee could run into the high six or seven figures and still represent a massive ROI for the developer. The company to beat isn't another SaaS startup. It's the calendar.
Sources
- [8 Clockwise, September 2026] 8 Clockwise Seed Fund Invests in Permeta.ai | https://www.linkedin.com/feed/update/urn:li:activity:7500946449011007489
- [Permeta.ai, Unknown] Company Website | https://www.permeta.ai/
- [Yahoo Finance/ACCESS Newswire, June 2026] ClassVI.AI Rebrands as Permeta, Expanding Its AI Platform For Energy Permitting | https://finance.yahoo.com/sectors/energy/articles/classvi-ai-rebrands-permeta-expanding-130000161.html
- [Startuply.vc, Unknown] Startuply.vc Profile | https://www.startuply.vc/
- [Caroline Soliman, LinkedIn, 2026] LinkedIn Post on Coastal Bend LNG Partnership | https://www.linkedin.com/in/caroline-soliman-361b3b69/
- [Carl Stumpf, LinkedIn, June 2026] LinkedIn Post on Pricing Model | https://www.linkedin.com/posts/carlstumpf_saas-isnt-coming-back-something-much-bigger-activity-7475260115751813120-m4eK
- [LinkedIn company profile, Unknown] Permeta.ai LinkedIn Profile
- [Permeta.ai About Us, Unknown] Team Background Page | https://www.permeta.ai/about-us