THE WATCHLIST

Today’s company is one you shouldn’t sleep on. You see a lot of content from me about the Ai Infrastructure space, and that’s because it’s hard to comprehend the growth potential. You don’t want to miss some of the the opportunities I’ll be highlighing over the next few months.

Apologies for the Share Link (bottom) not working last week - I was able to fix it, so if you’re enjoying this content, tell a friend and get some free digital goods on investing.

Let’s go…

Who is Nebius Group?

Nebius Group N.V. (Ticker NBIS) is a global AI infrastructure company based in Amsterdam. Think of them as a cloud provider built specifically for the AI boom. They design, build, and rent out the computing power that big companies and developers need to train and run artificial intelligence models.

NBIS is the successor of Yandex (Google of Russia). When Yandex separated its Russian and non-Russian operations because of international sanctions, NBIS took over the non-Russian side of the business; the data centers, AI infrastructure, and global cloud assets. They are led by Yandex’s former co-founder, Arkady Volozh, and has completely cut ties with its past Russian operations. Basically they kept the tech and talent, left the politics behind, and are now focused on building best-in-class AI infrastructure for the global market.

What Does NBIS Do?

NBIS builds and operates massive GPU data centers, basically the engine rooms that power modern AI.  They are vertically integrated, meaning they design, build, and operate almost every layer of infrastructure, instead of buying everything off the shelf or outsourcing it. This allows them to offer significantly lower prices than competitors.

Their goal is to become one of the leading “AI clouds” that provide high-performance computing for things like model training, inference, and large-scale data processing. Their cloud is designed for companies that need massive computing power for AI, but want it cheaper and quicker to access. They also build tools that help developers manage, deploy, and easily scale their AI workloads, so it’s not just hardware, it’s a full-stack AI platform.

On top of that, NBIS owns or has stakes in a few related tech businesses like ClickHouse (data analytics), Toloka (AI data labeling), Avride (autonomous driving tech), and TripleTen (tech education). These all tie back into their AI ecosystem.

💰️ How NBIS Makes Money?

1. Long-term infrastructure deals (60-65%)
Their biggest revenue source comes from large, multi-year contracts with companies that need serious GPU capacity. (Notable deal - they signed a $17 billion deal with Microsoft to provide AI infrastructure over several years)

2. Pay-as-you-go AI compute (20-25%)
Smaller customers and developers can pay to use their cloud platform by the hour, similar to AWS or Google Cloud, but optimized for AI workloads.

3. Software and tools (10%)
They offer developer tools, APIs, and managed AI services that sit on top of their infrastructure. These create additional recurring revenue streams.

4. Investments and partnerships (5%)
NBIS owns pieces of other fast-growing tech companies, which can add future upside and help drive demand for their own infrastructure.

Durable Moat

NBIS is building what they call a next-generation “AI cloud,” basically combining the hardware, software, and services needed to run large AI workloads all under one roof. That integration coule become a moat over time because it makes it harder for customers to switch once they’re fully embedded into their system.

The challenge is that the AI infrastructure game is very capital-intensive. Big players like AWS, Google, and Microsoft already dominate, so while NBIS has solid tech and a strong start, its advantage isn’t proven yet. They’re trying to stand out through efficiency, performance, and control over their full stack, but they still need scale, partnerships, and execution to turn that into a true moat.

Right now, I’d call their moat “developing”, not wide, but potentially getting stronger if they keep expanding and locking in long-term customers.

However 👇️ 

Nebius’s Developer Ecosystem and Integration Flexibility

Unlike most infrastructure providers, NBIS has built a fully unified platform that connects compute, storage, and AI dev tools in one interface, almost like a mini AWS, but purpose-built for AI. Developers can spin up training environments, deploy inference models, manage data, and monitor performance without juggling multiple vendors or tools. That level of integration dramatically improves developer retention and workload stickiness once a team sets up inside NBIS’s ecosystem, switching costs get high because their pipelines, datasets, and model configurations are fully embedded.

It’s not as flashy as a billion-dollar contract, but this “developer-first architecture creates long-term customer loyalty, the kind of moat that compounds over time.

Core Analysis

Market Position: In the GPU and AI cloud space, Nvidia is king. They own the hardware layer with nearly all the market share and power the entire ecosystem. Above that, AWS, Microsoft, and Google dominate large-scale AI cloud services. NBIS is a fast mover in the middle. They are a young GPU-focused, challenger building flexible, high-performance AI clouds. They’re growing fast and carving out a real niche; GPU-dense, AI-optimized infrastructure built purely for high-performance compute.  The stuff that needs extreme GPU power, low latency, and custom setups. 

Market Opportunity: Call it one of the fastest growing industries on the planet. The market for AI and GPU infrastructure is huge, from around $30 billion in 2023 to an estimated $250 billion-plus by 2030. NBIS is one of the fastest-moving players in that space. They are taking a broader, full-stack route, offering AI compute, storage, and tools across Europe and the U.S., and just landed a $17 billion partnership with Microsoft. They’re not a hyperscaler scale yet, but they are carving out serious market share.

Revenue Growth: YOY revenue growth has been huge….

2022→2023: +55%
2023→2024: +462%
2024→2025: est 450-500%

2025 est Revenue: $500-700m (some projections see them as high as $1b depending on the Microsoft deal.

Cash Flow: NBIS is spending aggressively to grab market share, but they are about efficiency and reach.  They are still cash flow negative, but running lean. They’re burning cash as they expand globally, but management says operating cash flow should be positive by late 2025 as their Microsoft deal ramps.

Profitability: They expect to reach positive adjusted EBITDA in the second half of this year, while full year adjusted EBITDA-positive by 2026.

Gross Margins: NBIS has been growing margins and benefiting nicely from AI infrastructure damand.

2022 = negative
2023 = 38%
2024 = 54%

Pricing Power: NBIS competes on affordability and flexibility, pricing 30–40% below hyperscalers to attract customers. NBIS’s advantage comes from cost efficiency, not premium pricing. Unlike a competitor like CoreWeave who is able to charge premium pricing (for now) due to scarcity.

Debt & Cash: Here’s a glimpse at their balance sheet. This business requires a good amount of leverage due to CapEx, however NBIS grows through partnerships and efficiency, where capital and infrastructure are shared rather than financed with debt (i.e. CoreWeave)

Cash: $1.68b
Long-Term Debt: $978m

Institutional Ownership: NBIS has about 44% institutional ownership

Why NBIS Looks Good 📈📈📈

• Massive Growth Runway: Revenue has exploded over the past two years and is still growing triple digits

• Big Partnerships: The $17 billion, multi-year deal with Microsoft gives them credibility, scale, and guaranteed revenue

• Cost Advantage: NBIS’s vertically integrated setup lets them build and run GPU data centers more efficiently, offering up to 40% lower prices than competitors

• Global Expansion: They’re already operating across Europe, the Middle East, and the U.S.

• Early Profitability Path: Management expects to hit positive EBITDA by late 2025, rare for an AI infrastructure company at this growth stage

Why NBIS Can Be Concerning 📉📉📉

• High Cash Burn: NBIS is still spending heavily on data centers, GPUs, and global expansion, maintaining negative cash flow

• Tough Competition: They’re up against AWS, Microsoft, and Google, plus CoreWeave — all with deeper pockets and established customers

• Execution Risk: The company’s scale-up is happening fast, and any delays in capacity buildouts could stump momentum

• Thin Margins (for now): NBIS competes on price, not premium positioning. That strategy attracts customers but limits pricing power

• Limited Track Record: As a relatively new public company, NBIS doesn’t have a long operating history

Scenario Analysis

Bull Case: NBIS becomes a serious player in AI infrastructure. The $17 billion Microsoft deal gives them scale, credibility, and momentum. By 2028, revenue’s in the billions, margins are strong, and they hit profitability by 2026. They build a name as the low-cost, high-performance alternative to the big cloud providers.

Base Case: NBIS keeps growing fast but slows down as competition ramps up. They reach profitability closer to 2028 with steady margins and a loyal customer base. The business matures into a solid niche player; efficient, global, and focused on AI, but not yet in the same league as the hyperscalers or Coreweave.

Bear Case:
Growth falls short of expectations. Costs stay high, profits stay negative, and market share slips to bigger competitors. By 2028, NBIS is still around, but as a smaller, price-driven provider fighting to stay relevant instead of leading the pack.

Zoom Out - 3+ Year Outlook

If Nebius executes, this company could grow into one of the major pillars of AI infrastructure outside the Big Three (AWS, Azure, Google). Over the next decade, the world’s demand for compute GPUs, AI training power, and inference workloads will grow exponentially. NBIS is positioning itself in the middle of that wave.

They already have the right blueprint: vertically integrated data centers, global footprint, and long-term partnerships like Microsoft. If they keep scaling efficiently, they could build a durable moat around low-cost, high-performance AI cloud infrastructure, becoming a key supplier for enterprises and governments needing localized, compliant AI compute.

By the early 2030s, NBIS could realistically be generating $10–15 billion + in annual revenue, with steady double-digit margins, transforming from a startup challenger into a profitable global operator.

The risk? Competition will be relentless. The hyperscalers won’t stand still, and GPU supply cycles will get tougher. But if NBIS continues to grow faster than the market, keeps costs low, and secures multi-year capacity deals, they’ll have a good shot at becoming the “next-generation cloud backbone” for AI-driven industries worldwide.

Bottom line: if AI infrastructure is the new oil, NBIS is early in the drilling phase, but they’ve already struck something big.