Working with Nik Storonsky, the founder and CEO of Europe’s most valuable startup Revolut, is relentless, says the man charged with overseeing his multi-million dollar VC fund, QuantumLight.
Since 2022, Ilya Kondrashov — an exited entrepreneur and Storonsky confidant — has been on the job of proving out the firm’s AI-led approach to investment. “The pace is monumental,” he tells Sifted. “He’s not going to let you relax before lighting a fire under you.”
In May, QuantumLight announced it had closed its first fund, raising $250m (of which Storonsky contributed more than $50m), as it looks to build out its proprietary deal-finding AI model. Whether that approach will win out remains to be seen, but the fund is staking its future on the tech’s ability to pick the right bets.
All 17 of the firm’s deals to date have been run through the algorithm, with no human overruling the model’s decision to invest — so far, at least.
QuantumLight has gained a reputation for taking a hands-on approach with the startups it backs, looking to instil the operating principles Storonsky used to scale Revolut in a series of “playbooks” it gives to founders.
And it’s not just portfolio companies working to Storonsky’s brief.
“Nik is very efficient,” says Kondrashov with a knowing smile when asked what it’s like to work with the Revolut founder. “He’s straight to the point — there’s a playbook for how meetings with him are structured.”
Working with Storonsky
The idea for QuantumLight first came about in 2017 over dinner when Kondrashov and Storonsky — the fund’s two cofounders and general partners — were trying to raise money for their then little known startups.
Back then Revolut, now valued at $45bn and having picked up $1.7bn in funding, had raised less than $20m. Kondrashov was cofounder and chief operating officer of B2B payments startup Kriya, which had raised a similar amount.
The two were frustrated by the “crowd mindset” of VCs, says Kondrashov. “No firm wanted to be the first to commit to a funding round, but when one committed, then suddenly many wanted to.”
The core thesis is that by using data and AI to make investment decisions, QuantumLight removes itself from the FOMO nature of VC.
It took another five years for the firm to launch, when Kondrashov had exited Kriya and Storonsky had more time to think about starting another business. Revolut has since become Europe’s most valuable startup, with more than 50m users and over $1bn in pre-tax profit in 2024.
While Kondrashov is running the day-to-day operations of the fund, Storonsky plays a part, not least as the public face of the firm, and his name and metrics of Revolut’s success are plastered all over the company’s website.
He holds something of a “president” role at the fund, says Kondrashov. The Revolut CEO helps to open doors to potential LPs and founders QuantumLight wants to invest in, he tells Sifted.
“[Storonsky is involved] in decisions like who we hire, how we position the company, who we choose as our LPs,” says Kondrashov.
“He makes decisions that don’t look obvious immediately,” says Kondrashov. “I can think ‘This is probably not how we should do it’, and then time passes and turns out he was dead right and I was wrong. It’s very detail oriented.”
Storonsky’s approach “can be quite contrarian”, he adds. “It’s quite different, but clearly it’s proved to be pretty effective.”
Running a fund with Storonsky also means working with his huge CEO office of around 30-40 people — who manage the Revolut founder's business interests and monitor performance of people who work for him, Kondrashov says.
AI-led dealmaking
As AI models grow more advanced, VCs are increasingly turning to the technology to outcompete rivals for deals. Sifted recently reported that around a third of investors now source at least half of their deals using these tools.
QuantumLight says it takes that approach to the next level — analysing 10bn data points and has tracked more than 700k VC-backed companies. While dataset size is not everything, targeting Series B and C startups the fund does have more of it to work with than other data-focused VCs like Moonfire which are looking to very early-stage signals.
Kondrashov claims that the AI model has made investment decisions on all of the 17 startups the fund has backed so far. The only reason Kondrashov and Storonsky, who do greenlight the deals, would overrule the model is if they found evidence of fraud during the due diligence process, he says.
The approach allows QuantumLight to “almost eliminate human judgement from the investment process,” says Kondrashov, adding that decisions can be made faster and are less likely to suffer from subjectivity and bias.
QuantumLight currently has a team of around 20, the majority of whom are engineers building out the firm’s AI model. Five are investors, carrying out due diligence, managing portfolio companies once deals are done — and are involved in the sourcing process.
“The model is producing a list of target companies ahead of their next round, then humans reach out and talk to founders of those companies and convince them to take our money,” Kondrashov says. “Then, when the next round happens the model makes the decision to invest — or not.”
While LPs other than Kondashov and Storonsky are undisclosed, they include a group of fellow billionaire tech founders, VCs and institutional investors, Kondrashov says.
Not everyone was convinced when the fund was out raising though. One LP that QuantumLight pitched to, who asked not to be named to protect business interests, tells Sifted that they didn’t invest because the approach wasn’t yet proven.
They also pointed out that other VCs — including EQT and Episode 1 — have been building deal-finding algorithms for longer and cautioned that the best deals at Series B and C will be oversubscribed and a tech platform won’t help win the deal.
The Revolut playbook
QuantumLight hopes having one of the darlings of European tech on-side will differentiate it from the competition.
“What fintech wouldn’t want Storonsky on their cap table?” said the LP, who didn’t invest.
It’s also selling itself on the prospect of portfolio founders being able to tap into the methods that took Revolut to becoming Europe’s standout tech success.
“One thing we did at the beginning is we sat down with Nik to understand how Revolut was able to grow so fast and beat well-funded competition by such a margin,” says Kondrashov.
QuantumLight has released a number of “playbooks” to its portfolio founders, which are operating manuals based on principles Storonsky used to scale Revolut.
It’s also released two of these publicly. One was on driving high performance among teams — which encouraged frequent performance reviews and said senior leaders should “act on low performance early” — and another on hiring strategy.
Problems with Europe
Though QuantumLight has a global focus and is headquartered in London, about 70% of the startups it’s backed are based in the US.
“We’re agnostic on industry and geography,” says Kondrashov — adding that the firm’s portfolio operate across AI, Web3, fintech, SaaS and marketplaces. “The model automatically picks companies that it thinks are most attractive, no matter where they arise.”
It’s a numbers game, he tells Sifted. There are just more startups to invest in in the US, so it stands to reason an AI model tasked with finding the best deals would find more that side of the Atlantic.
In Europe, QuantumLight has backed startups including British legaltech RobinAI and employee benefits startup Ben and Danish EV charging platform Monta.
Fragmentation is one thing holding Europe’s startups back, says Kondrashov. “Every country has different regulations, different languages and so expanding across that geography is harder than in the US, where everyone speaks the same language and regulations are more harmonious.”
Things can also move far slower this side of the Atlantic, he says.
“One thing I worry about are the transaction costs and speed in Europe,” says Kondrashov. “In the US we’re usually done doing a deal within one week, including wiring money.”
In Europe, however, deals can take far longer to go through, he adds. “We had a deal hanging for months because the company needed to have an in-person general meeting in front of a notary to vote to do some convertible note. I think that’s crazy.”
That won’t stop QuantumLight doing deals in the region though — so long as the model recommends it.
While the firm’s claim to be the first “truly systematic VC firm” is emblazoned across its website, the firm is staking its future on the performance of its tech unlike many before it.
“We don’t interfere with it using human decision making processes,” says Kondrashov. “The objective is to prove this approach delivers great returns.”
QuantumLight has backed about one company a month up until now, and will continue with that cadence, says Kondrashov. The firm will go back out to market for a second fund next year, he says — with, hopefully, more positive-looking data to show LPs.
