Soma Energy Raises $7M for AI Data Center Power Solutions

A new energy intelligence platform, Soma Energy, has officially launched from stealth, backed by a $7 million USD funding round. The Vancouver-based startup, founded by former Amazon Web Services (AWS) energy system architects, aims to tackle the escalating power demands of data centers through advanced AI-driven optimization of electricity infrastructure. This infusion of capital will fuel the expansion of their engineering and commercial teams, enabling broader deployment across North America.

The core of Soma Energy's offering lies in its sophisticated AI platform, designed to intelligently manage the interplay between power generation, battery storage, and the wider electrical grid. This orchestration is crucial as the digital infrastructure sector grapples with unprecedented energy consumption. Industry projections highlight the severity of the issue, with U.S. data center power demand anticipated to surpass the total electricity output of Texas by 2028, according to S&P Global. This surge presents a significant bottleneck for the rapid expansion of AI and cloud computing services.

The founding team, including CEO Ath Caramanolis, CTO Mario Souto, and Chief AI Scientist Henrique Hoeltgebaum, brings invaluable experience from their tenure at AWS. Caramanolis was instrumental in developing AWS's renewable energy optimization programs, while Souto led the creation of a machine learning platform managing gigawatts of renewable energy assets. Their firsthand understanding of managing power at hyperscale is the bedrock of Soma Energy's strategy to address grid constraints and interconnection delays that plague the industry.

“We witnessed and solved these challenges at AWS a decade ago – the grid limitations, the lengthy connection processes, the sheer complexity of managing power on a massive scale,” stated Caramanolis. “We established this company because we recognized a more efficient path forward.” Soma Energy emphasizes that the solution isn't merely building more power infrastructure, which can take five to ten years to materialize, but rather enhancing the intelligent utilization of existing resources. Their AI approach unlocks latent flexibility within current power systems, accelerating the availability of power for energy-intensive applications like AI.

The $7 million USD funding was secured across a pre-seed and a seed round, with the latter closing in Q2 2025. The seed round was spearheaded by Category Ventures, with significant contributions from Haystack, RRE Ventures, TO VC, and Uncork Capital. Pre-seed investors, including Panache Ventures and angel investor Walter Kortschak, also participated, underscoring strong market confidence in Soma Energy's vision. The company is already demonstrating traction, currently optimizing two gigawatts of electricity for power generation clients and engaging with five data center customers.

This development occurs amidst a broader industry push to modernize power grids for the AI era. The need for such solutions is underscored by recent comparable funding rounds, such as ThinkLabs' $28 million USD Series A, which also targets the challenge of preparing utility infrastructure for increased data center loads. Soma Energy's AI-powered optimization offers a critical pathway to unlock additional power capacity and reduce operational costs, positioning it as a key player in enabling the next wave of digital innovation without overwhelming existing energy resources.