The problem with grid-scale batteries is that you have to make money with them. It is a simple, brutal fact of physics and finance. A battery is a box of chemistry that can be charged and discharged, but the value of each electron depends on a market price that changes every five minutes across thousands of nodes. For a power company, managing a portfolio of these assets alongside wind, solar, and gas plants is a high-stakes, multi-dimensional puzzle. Equilibrium Energy, a San Francisco startup, is betting that the only way to solve it is with an AI system that can see the entire board.
Founded in 2021, the company has raised $71.7 million to date, including a $28 million Series B2 round in March 2025 led by GS Energy with participation from NRG Energy [Businesswire, March 2025]. Its flagship product, PowerOS, is an AI-native operating system for power companies, designed to forecast, optimize, and trade across a portfolio. The company calls it "connective intelligence for power" [Equilibrium Energy]. The pitch is straightforward: better predictions and faster decisions mean more money from the same assets, which is the only math that matters for decarbonization.
The AI wedge into the power stack
Equilibrium's wedge is not just better forecasting. It is the integration of that forecast into automated trading and risk management. The platform ingests data on weather, grid constraints, and market fundamentals, then uses machine learning to predict prices and recommend dispatch schedules for batteries and other flexible assets. For a developer with a 100-megawatt-hour battery in West Texas, the difference between a good and a bad algorithm can be millions in annual revenue. The company's early focus has been on this exact scenario, supporting grid-scale battery developers through tolling agreements and AI-powered trading strategies [Perplexity Sonar Pro Brief].
Its first publicly announced deployment is at NRG Energy, one of the largest competitive power producers in the United States [LinkedIn]. The strategic partnership with a leading global commodities trading house, announced separately, is a significant validation of its approach [Equilibrium Energy]. These are not science projects. They are commercial deals where the software's output translates directly into dollars.
From software to steel
The most concrete example of Equilibrium's model is a deal executed through Hatch EQ I LLC, a partnership with Hatch Renewables. The entity signed a tolling agreement with Jupiter Power for its 100-MWh Crossett I energy storage asset in West Texas [Perplexity Sonar Pro Brief]. In a tolling agreement, the software provider (in this case, Equilibrium through the partnership) takes on the market risk. It gets the right to dispatch the battery and collect the revenues, paying the asset owner a fixed fee. This structure proves the software's economic value in the most direct way possible: by putting its own balance sheet on the line.
- The portfolio layer. PowerOS offers application suites for portfolio management, construction, and development workbenches, aiming to be a central nervous system for power companies [Equilibrium Energy].
- The service layer. Core services include optimization, forecasting, risk management, and trading, wrapping around a company's existing data and systems [Equilibrium Energy].
- The commercial proof. The NRG deployment and the Jupiter Power tolling agreement move the company from PowerPoint to proven, revenue-generating operations.
This progression from SaaS to a capital-light tolling partnership shows a company trying to build a moat not just with code, but with commercial structures that are hard to replicate.
The team and the checkwriters
The founding team pairs commercial and technical expertise. CEO Ryan Hanley leads the company, while co-founder Jonathan Mather, who holds a PhD in Mechanical Engineering from UC Berkeley, serves as Head of Energy Science. Mather's background includes stints at Shell and Tesla [LinkedIn]. The investor list reads like a who's who of climate and deep tech conviction capital: DCVC, Breakthrough Energy Ventures, Valo Ventures, and Global Founders Capital are all backers [Perplexity Sonar Pro Brief]. The recent round added strategic muscle with GS Energy and NRG Energy, signaling that the company's customers are also willing to be its financiers.
| Role | Name | Key Background |
|---|---|---|
| Founder & CEO | Ryan Hanley | Not specified in sources |
| Co-Founder & Head of Energy Science | Jonathan Mather, PhD | PhD (UC Berkeley), Shell, Tesla |
The capital has funded a team of experts across energy, AI, software, and quantitative trading, a necessary blend for tackling the complexity of power markets [Equilibrium Energy].
Where the electrons could stop flowing
The risks here are not trivial. The power trading software space is crowded with incumbents like GE Vernova, Siemens, and a host of specialized boutiques. These are well-funded, entrenched players with decades of domain knowledge. Equilibrium's answer is its AI-native, integrated approach and its willingness to engage in tolling agreements, which many software vendors avoid. Still, selling into conservative, regulated utilities is a long and expensive game. A single failed forecast during a grid emergency could erode years of trust.
The company also faces the classic innovator's dilemma of a startup: it needs to prove its platform can scale from managing a few batteries to orchestrating entire, diversified power portfolios. The NRG deployment is a critical first step, but it is just one step. The next twelve months will likely be about landing a second major utility customer and proving that the revenue uplift from PowerOS consistently beats the legacy benchmarks.
The math on a megawatt-hour
Consider the Jupiter Power battery in Texas: 100 megawatt-hours of capacity. In the volatile ERCOT market, the spread between the lowest and highest price in a day can regularly exceed $100 per megawatt-hour. A human trader might capture a fraction of that. An AI system tuned to forecast those spikes could, in theory, capture more. If Equilibrium's software can improve the battery's annual revenue by just 10% over a baseline,a plausible claim for a sophisticated optimizer,that could mean an extra several hundred thousand dollars per year on a single asset. Scale that across a portfolio, and the unit economics start to justify the software's cost. The company's reported annual revenue of $69.6 million in 2026 suggests it is already translating these calculations into a business [RocketReach].
For Equilibrium to become the default operating system for power companies, it must beat the incumbent giants on the only metric that matters: dollars per megawatt-hour. Its early deals suggest it has a shot, but the real test is whether it can move from the battery edge to the thermal heart of the grid.
Sources
- [Businesswire, March 2025] Equilibrium Energy Raises $28M to Unlock Full Potential in Power Portfolio Management | https://www.businesswire.com/news/home/20250312583176/en/Equilibrium-Energy-Raises-$28M-to-Unlock-Full-Potential-in-Power-Portfolio-Management-and-Launches-its-Flagship-EQ-Mission-Control-Software-Platform
- [Equilibrium Energy] Connective Intelligence for Power | https://www.equilibriumenergy.com/
- [Perplexity Sonar Pro Brief, October 2024] Company overview and funding details |
- [LinkedIn] Company page and deployment announcement | https://www.linkedin.com/company/equilibrium-energy-inc
- [RocketReach] Annual revenue figure |
- [Equilibrium Energy] Strategic partnership announcement | https://www.equilibriumenergy.com/news/equilibrium-energy-announces-strategic-partnership-with-leading-global-commodities-trading-house-to-develop-ai-native-power-trading-strategies