Last month Paul Graham met with a little known nuclear startup.

Apollo Atomics went through YC's Spring 2026 batch and recently announced its seed round.

It is about as early as a company can get, and USVC investors get exposure to it.

That may surprise some of you reading.

We’ve written a lot lately about SpaceX, Anthropic, and Crusoe. Companies with real revenue and IPOs on the calendar. That's one side of the fund.

This is about the other side: companies at the earliest stage, sometimes just months past their earliest financing.

Early companies can and often do fail, but the ones that don't have room to grow.

Here are three early-stage companies that came to USVC through private fund managers we invested in (more on that below):

TLDR: Nuclear reactors built in a factory and delivered on a truck.

Nuclear plants can take a decade and up to $17 billion to build on site. Most advanced reactor startups try to fix this by reinventing everything: new fuel, new coolant, new regulatory path.

Founder Assil Halimi did the opposite. He kept the proven pressurized water reactor, and changed a single part instead: the steam generator. This is the biggest component in a plant and was borrowed from coal designs in the 1950s. Apollo reportedly made it 10x to 20x smaller so it could fit inside a factory.

Halimi did his PhD on this at MIT and ran fuel reloads for Belgium's nuclear fleet before starting the company. His thesis: size is cost, power is revenue.

The numbers: Target of 3 cents per kilowatt hour. A 300 MW plant in under 24 months. A working 40 kW test reactor at MIT. 20 GW of signed letters of intent. And its seed was led by FCVC with participation from YC, Stanford, and MIT's E14 Fund.

Our take: Hyperscalers are waiting years to connect new data centers to the power grid. A factory-built reactor that sits on site, powering the data center directly is the product that the market has been asking for.

TLDR: A school in San Francisco for teenagers who'd rather build than take tests.

CEO Gagan Biyani co-founded two well-known edtech startups of the last two decades, Udemy and Maven. His argument with this venture: the best AI researchers left universities for companies, so he's building a school with those companies as partners. Students will spend most of their time on their own projects and classes exist only to feed those projects.

a16z incubated the startup and led a $42 million round. Marc Andreessen and Erik Torenberg sit on the board and they’ve launched with ten founding partners: Anduril, Anthropic, Coinbase, Google, Meta, NVIDIA, OpenAI, Palantir, Replit, and Stripe.

The numbers: First class of about 50, tuition-free, fall 2027. A paid two-year program follows in 2028, pending regulatory approval, with tuition estimates at $60,000 to $90,000 a year.

Biyani is clear that this is meant to be a business, not a charity. Three revenue lines drive the business: tuition, corporate partnerships, and equity in the companies students start. And their recently announced fundraise was sized for years of runway rather than 18 months.

Our take: Back in May, we argued that AI has made it far cheaper to build, and that the number of ambitious founders is growing faster than the money looking for them. The Academy is a bet on finding them even earlier.

TLDR: A bank account your AI agent can spend from but can't drain.

This is likely the least known company of the three, and yet, potentially the most timely.

In the past month, AI agents started acting on their own. Instinct, a personal agent recently valued at $10 billion, now creates its own email accounts, signs up for services, and pays for things through Stripe. Meta's Muse passed ChatGPT's early mobile numbers in 12 days, per Apptopia estimates.

An agent that books your trip needs to pay for it. An agent that runs your side business needs to invoice, get paid, and pay contractors. And just handing it your credit card is not a plan.

Perflo's answer? Real accounts with delegated access. Agents get their own bank account and can pay from it by card, x402, wire, or ACH. Every payment runs inside limits set by the user: per-payment caps, rolling daily and monthly limits, allowlisted payees. Every payment is checked against the rules before it moves, and when a payment is refused the agent gets told why so it can ask a better question.

The numbers: The company says it's live in 80+ countries with accounts in nine currencies, settling over ACH, SEPA, Pix, and UPI. It operates as a registered money services business through regulated partners.

Our take: When we invested in a private vehicle that holds Mercury, we argued that every so often a basic layer of business infrastructure gets rebuilt around the way companies really operate. If AI agents start running parts of those companies, they'll need infrastructure built for them too.

By the time a company's name reaches your brokerage account, most of its value may have already been created. Anthropic, for example, was worth about $4 billion in early 2023. Its last preferred round, in May, valued it at $965 billion.

There are two ways one may get in before an IPO. One is buying a late-stage company while it's still private. You get some of the growth, but the purchase price already marks what investors know.

The other can be buying into an early-stage company. You may often pay less to get in, but most of these companies can fail, and it can take a decade to find out which ones won't.

Venture returns generally follow a power law. This argues that a handful of companies can drive most of the results, and it can take years to know which ones. In other words, a company that can look like a modest win in year three can turn into a far bigger one by year eight.

But for most individuals, both paths to investment have been out of reach. Late-stage rounds and early-stage funds have generally required accreditation or high minimums.

USVC aims to give individuals exposure to both, with no accreditation requirement.

We invest across stages in three ways: late-stage secondaries to get exposure to companies you may recognize, buyouts of LP stakes in funds a few years old, and fresh commitments to early-stage private fund managers investing in the current vintage.

We didn't source Apollo Atomics, price it, or hear the pitch. We backed a private fund that did.

We believe effective early-stage managers spend days with companies being formed right now. They may see the batch before demo day and get the intro before the round is announced. When we commit to one of them, their access becomes ours indirectly.

That's how an investment in USVC reached exposure to a company that raised its seed round only last month.

USVC is a registered fund that helps broaden access to venture capital.

Invest from $500 at usvc.com, any business day, at NAV.

If you already own USVC, you already have exposure to Apollo Atomics, the Horowitz Andreessen Academy, and Perflo. Alongside Anthropic, Supabase, Mercury, and the rest.

USVC invests in portfolio companies through private funds and special purpose vehicles. For important information and USVC holdings, please see https://usvc.com/portfolio

— Ankur

P.S. If this email got you thinking about the next wave of founders, come build alongside them.

Andrew Yeung and I are hosting the Outliers Summit in NYC on November 5 & 6. The first day will focus on tactical sessions from experts in the disciplines that actually matter in the AI era. The second will include candid talks and fireside chats with founders like Sam Parr, Sahil Lavingia, and Ryan Hoover.

Plus, every ticket holder now gets access to $100,000+ in credits to what we believe to be the best tools for creating, building, and shipping.

That includes products from Vercel, ElevenLabs, Granola, Datadog, Fin, Stan, Fyxer, Thoropass, Zo Computer, Agree.com, Flora, Digital Executive, Zypsy, and Airtaskr, with more to be announced.

Two-day passes go from $199 to $219 this Wednesday, September 30, so lock in your spot now!

P.P.S. Q3 wraps up this week and I'm hosting a free, live recap on Wednesday October 7th @ 2pm ET.

Bring your questions for the live Q&A, or register and we'll send you the replay.

Investors should carefully consider the investment objectives, risks, sales charges and expenses of USVC before investing. USVC's prospectus contains this and other information and may be obtained at http://usvc.com/prospectus or by calling +1 (888) 200-4361. Read the prospectus carefully before investing.

This communication is for informational purposes only, is not intended to be a recommendation for any investment or other advice of any kind and shall not constitute or imply any offer to purchase, sell or hold any security or to enter into or engage in any type of transaction. Any such offers will only be made pursuant to USVC's prospectus, which should be carefully reviewed before investing.

Investing in the USVC Venture Capital Access Fund involves significant risk, including the possible loss of principal. Venture capital investments are speculative, illiquid, and subject to a high degree of risk. Past performance does not guarantee future results.

USVC Venture Capital Access Fund is distributed by North Capital Private Securities Corporation (NCPS), member FINRA/SIPC. NCPS is not affiliated with USVC's adviser or its affiliates.

Investing in USVC's shares involves substantial risk, including the potential loss of your entire investment. Shares are not listed on any exchange, are illiquid, and liquidity is limited to periodic repurchases at the discretion of the Board, which are not guaranteed. This investment is speculative and suitable only for long-term investors who can bear the risks of limited liquidity. Certain conflicts of interest involving USVC and its affiliates could impact USVC's investment returns and limit the flexibility of its investment policies. Past performance does not guarantee future results. Fees, expenses, and conflicts of interest may reduce returns.

USVC's shares have no history of public trading. You should not expect to be able to sell your shares other than through USVC's repurchase policy, regardless of how USVC performs. USVC does not intend to list its shares on any securities exchange during the continuous offering, and it does not expect a secondary market in the shares to develop.

USVC invests in portfolio companies directly, through funds, and through special purpose vehicles funds which are subject to certain risks including those related to illiquidity, indirect fees, valuation, limited operating histories and limited information regarding underlying investments. As a result of the foregoing, an investment in USVC's shares is not suitable for investors that require liquidity, other than liquidity provided through USVC's repurchase policy. The amount of distributions that USVC may pay, if any, is uncertain.