New flexible interconnection frameworks promise faster timelines for utility-to-datacenter power connections, reducing project lead times.

Hyperscalers racing to bring AI data centers online frequently encounter a critical bottleneck: connecting to the U.S. power grid is rarely straightforward. The electric infrastructure required to support compute-intensive projects must often be built or upgraded from scratch—a process that can take years. Many developers opt for on-site, behind-the-meter generation solutions in the interim, but most prefer utility service and prefer it as soon as possible.

That preference is precisely what has captured Google's attention, particularly in the context of what it views as a "generational opportunity for AI to redesign a sustainable and resilient global energy system." Google recently selected 28 energy-focused, AI-driven startups for its 10-week Google for Startups Accelerator program, running from September 8 to November 12, 2026. Camus Energy, a California-based company, was among a dozen North American firms selected for the initiative. The company's FlexConnect software promises to reduce typical interconnection wait times by three to five years by utilizing flexible operating limits within existing infrastructure.

Google's accelerator program aims to help AI data centers connect to the grid faster through new energy transition techniques that leverage existing poles and wires. Participants receive technical workshops, strategic consultations, and two in-person milestones with Google's network of utility and investment partners. The curriculum emphasizes grid modernization, demand flexibility, and clean tech infrastructure, enabling startups to develop solutions spanning electricity generation to grid operations.

As part of the program, Camus Energy will accelerate FlexConnect's development and demonstrate it to major utilities including Duke Energy, Edison International, and PG&E. The demonstration will provide a practical blueprint for interconnecting data centers using flexible operating limits to protect grid reliability during utility upgrades.

"The program gives hyperscalers a chance to move past the all-or-nothing approach to grid access, proving it is possible to reliably unlock massive new capacity while keeping the local grid stable, clean, and affordable," said Astrid Atkinson, CEO and co-founder of Camus Energy.

David Bailey, a veteran software engineer with more than two decades at Google and Amazon where he built large-scale search and machine learning systems, now applies that expertise at Camus Energy, transforming complex grid data into reliable, scalable intelligence. Bailey responded to questions following Camus's selection by Google; the following is an edited transcript of that exchange.

**On flexibility and data centers:**

There is considerable value in flexibility, and the elegant part is that everyone benefits. The data center connects more quickly because a bottleneck is removed. The utility gains a substantial customer. A little flexibility goes a long way—the data center receives nearly all the power it originally needed since curtailment is typically rare. By solving this through software, costly grid upgrades are avoided, potentially making electricity more affordable for the average citizen. Additionally, by better utilizing existing and increasingly green grid capacity rather than building inefficient off-grid gas plants, flexibility supports climate goals.

**How FlexConnect works technically:**

The mechanism is straightforward. The data center connects to the grid with a commitment comprising firm power plus conditional power that is not guaranteed. FlexConnect cooperates with the utility to monitor and analyze grid conditions combined with weather forecasts to predict the rare moments when conditional power will be unavailable. The platform communicates an "operating envelope" to the data center with advance notice when possible, allowing them to plan responses using battery backup, job deferral, or other means.

**On the Google accelerator selection:**

I was thrilled when Camus was selected—it represents a significant opportunity to deepen our relationship with Google. We are already collaborating with Google to deploy FlexConnect in support of their data center expansion. FlexConnect itself runs on Google Cloud, and we can benefit from enhanced expertise with the platform, especially as its AI capabilities expand. As a former Google engineer who worked on machine learning algorithms for Google Search and their implementation across Google's compute fleet, returning to that partnership feels particularly meaningful.

Regarding fellow selected companies, I view them through a collaborative rather than competitive lens. We form a small community and need all hands on deck to address these critical societal challenges. I hope we can strengthen each other. I have long admired Grid Status and their sophisticated visualizations, and I am equally keen to meet companies new to me.

**What utilities will see from demonstrations:**

We aim to demonstrate that flexible interconnection works in practice. We possess the tools and technology to execute it. Software-based orchestration of flexible assets can serve as a first-class, reliable solution for grid stability.

Camus has operated in this space for several years, initially focusing deeply on the utility side, building significant expertise and trust in integrating utility data into dashboards and insights operators appreciate. Over the past year, we have expanded to data center applications with promising early studies across multiple developers. While this aspect is newer, it leverages our past experience, particularly Astrid Atkinson's pioneering work on compute-side orchestration that powers modern data centers. It remains early in proving this connection between the two worlds, but we are committed to making it succeed.

**Positioning and the future of speed-to-power:**

Our unique strength lies in utility experience. Years of hands-on analytics for operators and planners, combined with coordination of smaller flexible assets, position us as a trustworthy partner in ensuring flexible connections genuinely improve grid reliability rather than undermine it. Equally important, we understand intuitively what hyperscalers require from their data centers to meet their own reliable compute commitments and the ways flexible power can accommodate those needs.

---

**Other North American startups selected for the Google accelerator program:**

Anode (United States): AI-orchestrated virtual utility mobilizing microgrids on demand

Cosmic Robotics (United States): autonomous construction equipment for solar farms and data centers

Dig Energy (United States): modular drilling technology enabling geothermal heating, cooling, and storage

Felix Fusion (United States): AI-powered platform helping data centers, mining operations, and industrial projects find energy and connect to the power grid flexibly

Fram Energy (United States): revenue collection layer for local, independent energy projects

Grid Status (United States): modern data and market intelligence platform for the energy industry

Hestia Energy (United States): solution finding and unlocking geologic hydrogen reservoirs

Nyquis (United States): solution cutting time to power for data centers and large new loads

Optiwatt (United States): electric vehicle–first energy platform connecting vehicle brands with programs balancing energy usage and demand across North America

Swish Solar (Canada): operating system for the world's solar infrastructure

Zola Intelligence (United States): AI-powered platform embedding automation and AI agents into finance and operational workflows with smart hardware for energy delivery