Helix Is Coordinating the Physical Stack Behind AI

AI can scale in software faster than the physical world can pour concrete, connect a substation, or bring a transmission line online. That gap is becoming one of the defining constraints of the AI economy.

Helix Digital Infrastructure was created to work on that gap as one company.

Launched in June 2026, Helix is designed to finance, develop, operate, and manage the infrastructure hyperscalers need to bring new AI capacity online. Its scope crosses hyperscale data centers, power generation and transmission, fiber, and connectivity. The company is not selling one more component into the stack. It is trying to coordinate the stack itself.

That distinction matters. A data center project can have capital and still wait on power. It can secure power and still face interconnection, permitting, network, or construction delays. Every handoff between those systems adds another contract, timeline, and failure point.

Helix is betting that the winner will not simply own assets. It will make a fragmented buildout behave like one operating system.

What Helix Digital Infrastructure Does

Helix positions itself as a single strategic partner for large AI infrastructure users. Instead of asking a hyperscaler to coordinate data-center development, power procurement, grid access, connectivity, and long-duration financing across separate providers, Helix wants to manage those workstreams through one platform.

The company launched with more than $10 billion in committed capital from founding investors KKR, Kuwait Investment Authority, NVIDIA, and Vistra. That structure brings together several capabilities that usually sit in different rooms.

KKR contributes infrastructure capital and investment experience. NVIDIA is a strategic technology partner around AI factory-aligned infrastructure. Vistra is Helix's preferred power provider, bringing generation and power-market expertise. The Kuwait Investment Authority adds long-duration institutional capital.

Samsung then added another $1 billion strategic commitment. Samsung Electronics committed $500 million, while five other Samsung affiliates supplied the remaining $500 million.

The capital is real. So is the strategic backing. But it is important to call it what it is: evidence of financial capacity and ecosystem alignment, not proof of customer revenue, completed projects, or deployed megawatts.

The Coordination Problem Behind AI Infrastructure

AI infrastructure is often discussed as a race for chips. The physical bottleneck is broader.

Large deployments require suitable land, data-center design, electrical generation, transmission, grid interconnection, network capacity, equipment, construction labor, permits, and contracts that can survive a long operating life. Each dependency moves on a different clock.

That is why Helix's model is more interesting than a large capital pool by itself. The company is trying to reduce the transaction cost between those dependencies.

Its partner architecture is designed around that objective. NVIDIA provides a view into AI factory requirements. Vistra can connect development plans with real power-market and generation capability. KKR can underwrite projects with patient capital and draw on an infrastructure platform spanning digital and power assets.

If the model works, hyperscalers gain one partner accountable for coordinating more of the path from demand to operating capacity. That could shorten decisions, make projects repeatable, and reduce the risk that one missing input stalls the entire system.

The hard part is execution. No amount of capital makes interconnection queues disappear. Transmission upgrades, permitting, equipment supply, community approval, and construction still have to happen in the real world.

Leadership Built Around Cloud and Infrastructure

Helix is led by Adam Selipsky, its Co-Founder and CEO. Selipsky previously led Amazon Web Services and Tableau, giving him a customer-side view of how cloud platforms buy, scale, and operate infrastructure.

Waldemar Szlezak serves as CIO while continuing as KKR's Global Head of Digital Infrastructure. His investment work spans data centers, fiber, broadband, and communications infrastructure across multiple markets.

In September, Helix added two more operators to its leadership team. Jonathan Lin and Kleber Costa joined to lead data-center and energy efforts, respectively. Lin brings more than a decade of Equinix experience across strategy, operations, growth, and hyperscale customers. Costa most recently served as Chief Commercial Officer of AES Clean Energy and U.S. Generation.

Those additions are an early signal about how Helix intends to build. The team is not organized around a single asset class. It is assembling leadership across cloud demand, investment, data-center delivery, and energy.

What the Market Signal Actually Says

Helix has no long public operating history. It launched in June. There is no verified public customer list, revenue figure, completed-project count, or deployed-capacity metric.

That means the strongest evidence today sits upstream of delivery: more than $11 billion in announced commitments, strategic partners with technology and power capability, and leaders who have operated cloud, data-center, energy, and infrastructure-investment platforms.

The absence of a public careers page reinforces how early the company is. Helix's current team buildout is visible through leadership additions, but unrelated job pages for similarly named companies should not be treated as Helix Digital Infrastructure openings.

For operators and investors, the next meaningful milestones will be physical. Which projects does Helix back? How quickly can it move them through power, permitting, and construction? Can an integrated model outperform specialists without becoming another layer of coordination?

Why Helix Matters Now

The AI buildout is forcing software companies to become students of energy systems, real estate, project finance, and industrial delivery. Helix is trying to turn that complexity into a service.

Its thesis is straightforward: hyperscalers do not need another vendor selling one piece of the problem. They need a counterparty capable of coordinating capital, compute infrastructure, power, and connectivity over the same time horizon.

The company has enough capital and partner strength to make that thesis credible. Now it has to make it repeatable.

That is the real Helix test. Not whether it can announce another commitment, but whether it can turn a coordinated balance sheet into energized, connected infrastructure faster than the fragmented system it was built to replace.