top of page
4Artboard 3_2x_edited_edited.png

The UK’s home for tokenised equity. Independent news, insight and resources for founders raising capital, investors deploying it, and the firms supporting both — as the regulation, infrastructure and opportunity converge.

QuantumLight Closes 500 Million Dollar Second Fund

  • Writer: Shawn Jhanji
    Shawn Jhanji
  • Aug 18
  • 3 min read
QuantumLight, the venture firm founded by Revolut chief executive Nik Storonsky, has closed 500 million dollars for its second fund, doubling the size of the 250 million dollar debut vehicle it closed just over a year ago. It appears QuantumLight is trying to prove that an early stage investment decision can be made by a machine reading millions of data points rather than a partner reading a pitch deck and a warm introduction. What happened QuantumLight was founded in 2022 by Storonsky alongside chief executive Ilya Kondrashov. Its central proposition is Aleph, a proprietary AI system that the firm says has evaluated more than 700,000 venture backed companies and has recommended all 17 of the deals QuantumLight has made to date, including UK legal AI company Robin AI and employee benefits platform Ben. The firm's pitch to limited partners is that removing human judgement from sourcing and screening reduces the pattern matching and network bias that shapes which founders get seen in the first place. The first fund, which closed at 250 million dollars in May 2025, was backed by a mix of undisclosed institutions and tech founders. The second fund, which had been targeted at 500 million dollars since talks with LPs began in January, has now closed at that target, giving QuantumLight roughly 750 million dollars in total firepower two years into its life. What it means for founders The UK funding gap story is usually told through who gets the cheque: female founders raise a fraction of the pool, founders outside London struggle for warm introductions, and pattern matching by partners tends to reproduce the networks those partners already have. QuantumLight's bet is that an algorithmic first pass, built on hard performance data rather than who a founder knows, can widen the aperture of who gets looked at at all.

QuantumLight, the venture firm founded by Revolut chief executive Nik Storonsky, has closed 500 million dollars for its second fund, doubling the size of the 250 million dollar debut vehicle it closed just over a year ago.


It appears QuantumLight is trying to prove that an early stage investment decision can be made by a machine reading millions of data points rather than a partner reading a pitch deck and a warm introduction.


What happened.


QuantumLight was founded in 2022 by Storonsky alongside chief executive Ilya Kondrashov. Its central proposition is Aleph, a proprietary AI system that the firm says has evaluated more than 700,000 venture backed companies and has recommended all 17 of the deals QuantumLight has made to date, including UK legal AI company Robin AI and employee benefits platform Ben.


The firm's pitch to limited partners is that removing human judgement from sourcing and screening reduces the pattern matching and network bias that shapes which founders get seen in the first place.


The first fund, which closed at 250 million dollars in May 2025, was backed by a mix of undisclosed institutions and tech founders. The second fund, which had been targeted at 500 million dollars since talks with LPs began in January, has now closed at that target, giving QuantumLight roughly 750 million dollars in total firepower two years into its life.


What it means for founders


The UK funding gap story is usually told through who gets the cheque: female founders raise a fraction of the pool, founders outside London struggle for warm introductions, and pattern matching by partners tends to reproduce the networks those partners already have. QuantumLight's bet is that an algorithmic first pass, built on hard performance data rather than who a founder knows, can widen the aperture of who gets looked at at all.


It is a genuinely interesting attempt at structural reform, and it deserves scrutiny rather than either dismissal or uncritical applause. An AI model trained on the historical record of which companies succeeded is also a model trained on a market that has, for decades, underfunded women, ethnic minority founders and those outside London and the golden triangle. Removing a human partner's bias does not automatically remove the bias baked into the data that trained the machine. QuantumLight has not published detailed figures on the diversity of its portfolio, and founders and commentators are right to ask for them as the firm scales toward three quarters of a billion dollars under management.


What is genuinely promising is the principle QuantumLight is testing at meaningful scale: that sourcing does not have to run through a partner's contact book. If systematic, data first sourcing can be shown over time to surface founders that warm introduction networks miss, it becomes a template other funds can borrow from, alongside blind pitch processes, open application windows and the scout networks already expanding access across the UK.


What comes next


QuantumLight has said little publicly about how it will deploy the new capital beyond continuing to back growth stage companies identified by Aleph. The more interesting test for the wider ecosystem will be whether QuantumLight, or funds like it, start publishing outcome data on who Aleph surfaces and backs, so the claim that algorithmic sourcing reduces bias can be checked against evidence rather than taken on faith.


Key Takeaways


  • QuantumLight has closed 500 million dollars for its second fund, taking the firm's total capital to roughly 750 million dollars since its 2022 founding.

  • The firm's AI model, Aleph, has recommended all 17 of its deals to date, including UK companies Robin AI and Ben, as part of an explicit bet that algorithmic sourcing reduces the human pattern matching bias that shapes access to capital.

  • The approach is a genuine attempt at selection reform, but AI models trained on historical success data can also encode the same structural bias they aim to remove, making outcome transparency the real test.

  • For UK founders, the story is less about one firm's fund size and more about whether data first sourcing becomes a credible, evidenced alternative to warm introduction networks.


Sources:

Comments

bottom of page