We value your privacy

    We use cookies to analyze our traffic. By clicking “Accept all”, you consent to our use of cookies. Privacy notice

    Skip to content
    CrossWork
    Request access
    Illustration: an aerial view of a data-center campus

    Strategy

    Where we look within AI: the infrastructure it runs on, and AI delivered as a service.

    Our focus is top late-stage companies in any sector. Within AI we look for the physical infrastructure every system depends on, and for companies that sell AI's results as a service.

    Market map

    The infrastructure AI runs on, and AI delivered as a service

    Where we look within AI, layer by layer.

    Chips and compute

    • Purpose-built inference silicon is collapsing the cost per token.
    • Every task an AI system takes on runs on this layer, so demand compounds.

    Power and cooling

    • AI data centers draw far more power per rack than the facilities built before them.
    • Heat has to be removed at the chip: liquid cooling is moving from niche to standard.

    Data centers and networks

    • The buildings, fiber and switching that connect compute to its users.
    • Capacity is planned years ahead, which favours operators with land, power and customers secured.

    AI delivered as a service

    • Companies that sell the finished result to the end customer, not a software tool for someone else to run.
    • Priced on the work done, so the customer buys an outcome rather than a licence.
    Illustration: an electrical substation feeding a data center
    Illustration: bundled fiber-optic cables
    Illustration: a liquid-cooling manifold between server racks

    Thesis

    1. 01

      The value moved before the IPO

      The last platform shifts rewarded public-market investors because companies listed early. This one does not. The companies building AI infrastructure and models are staying private for longer and raising more privately than any generation before them.

      The consequence is simple: the steepest part of the value curve now sits before the IPO.

    2. 02

      Access is relationship-gated

      The companies that matter choose their investors. Founders choose who gets in, and they choose people who can help.

      By the time a company is broadly available, the repricing has happened.

    3. 03

      Operators see the winners first

      The people who can tell a platform from a product are the people who have run the functions it replaces. That is why CrossWork is built around an advisory board of former operators.

      They see the companies first, they sit in diligence, and their introductions open doors.

    The window

    Platform shifts have short pre-IPO windows. This is one of them.

    Each platform shift produced a small set of category leaders, and once they listed the private window on them closed. The pattern is repeating now, faster.

    • Cloud · 2012

      ServiceNow, a cloud category leader, listed in 2012.

    • Mobile · 2012 to 2013

      Facebook listed in May 2012 and Twitter in November 2013: the social-mobile leaders went public within eighteen months of each other.

    • Now · 2026

      Cerebras listed in May 2026 and SpaceX in June 2026. OpenAI and Anthropic have filed confidentially for IPOs, and Discord reportedly filed confidentially in January 2026.

    Sources
    1. 9. SEC EDGAR, 2012: ServiceNow, Inc. prospectus (Form 424B4).
    2. 10. Facebook Newsroom, May 17, 2012: Facebook announces pricing of initial public offering; shares expected to begin trading on the Nasdaq Global Select Market on May 18, 2012.
    3. 8. SEC EDGAR, November 6, 2013: Twitter, Inc. prospectus (Form 424B4).
    4. 1. CNBC, May 13, 2026: Cerebras prices IPO above expected range.
    5. 6. SpaceX, June 11, 2026: pricing of initial public offering (Nasdaq: SPCX).
    6. 5. TechCrunch, June 8, 2026: Following Anthropic, OpenAI files confidentially for IPO.
    7. 7. Bloomberg, January 6, 2026: Discord is said to file confidentially for IPO.

    Market

    How large, by others' estimates

    Global AI token consumption is expected to grow about 40 times over the next four years, according to Qualcomm CEO Cristiano Amon (COMPUTEX keynote, June 1, 2026).

    MarketsandMarkets sizes the AI-agents market at about $8 billion in 2025, rising to roughly $53 billion by 2030, about 46% a year.

    PwC estimates generative and agentic AI could add US$2.6 to 4.4 trillion a year to global GDP by 2030.

    Sources
    1. 2. Seeking Alpha, June 2026: Qualcomm CEO Cristiano Amon sees a 40-fold AI token surge by 2030 (COMPUTEX 2026 keynote, June 1, 2026).
    2. 3. MarketsandMarkets, 2025: AI Agents Market worth $52.62 billion by 2030.
    3. 4. PwC Middle East, 2024: Agentic AI, the new frontier in GenAI.

    Discipline

    Real companies. Real revenue.

    What we buy, and how we buy it.

    We look for companies with real revenue, real customers and a visible path to liquidity, bought at a large discount to the last round.

    Discipline on revenue and price is built into every position.

    Roughly $100M+ in annual revenue and a $1B+ valuation, with a visible path to liquidity.

    Top late-stage companiesValued above one billion dollarsAbove one hundred million dollars of revenueA visible path to liquidityBought at a large discount,through the advisory board
    1. Top late-stage companies
    2. Valued above one billion dollars
    3. Above one hundred million dollars of revenue
    4. A visible path to liquidity
    5. Bought at a large discount, through the advisory board
    Illustrative geometry. Criteria as stated above.

    Built for every cycle

    • Real revenue only.
    • We also hold non-AI businesses that benefit from AI's cost savings.

    The window is now.

    Request accessLogin