Brighteye Places a $72 Million Bet on Human Capability

Brighteye Ventures has taken a $72 million first close on Fund III, lifting the firm to $245 million in assets. The Luxembourg-domiciled investor, which works from London and Paris, is not stretching the vehicle. Fund III is aimed at €100 million, about $115 million, the same hard cap as Fund II.

Founding partner Benoit Wirz is using that ceiling to buy more early shots at what the firm now calls HumanOS: software that helps people learn, work and adapt as models get cheaper. The first cheques already sit in a US classroom layer, a hospital infection tool, and an Italian electrician marketplace.

Brighteye’s $72 Million First Close Holds Fund Size Steady

Alex Spiro Latsis and Wirz started Brighteye in 2017 to back early European companies in learning and work. The third fund still lives in that market. The construction has shifted. Wirz said the book is “skewed slightly earlier than it did before because I think you can build more with less,” and that early traction in this cycle “tends to accelerate much faster.”

FUND III AT FIRST CLOSE

• First close: $72 million, announced 16 September 2026.
• Assets: $245 million across the firm.
• Target: €100 million, about $115 million, with a final close planned for the first half of 2027.
• Pace: up to 35 investments across AI, learning, productivity and labour infrastructure.

More than 90 percent of the capital is reserved for core cheques of $0.5 million to $4 million. The rest is held for “blink” tickets of $150,000 to $500,000 at idea stage. That is a lot of names for a fund that is not larger than the last one. It is the wager: more companies, earlier, in the same envelope.

Since 2017 the firm has backed more than 50 companies. Those holdings have gone on to raise over $1 billion in later rounds, and several now generate more than $100 million in annual revenue. Returning limited partners include the European Investment Fund and the Jacobs Foundation, alongside European family offices. New names on the first close are Lumina Foundation, Zanichelli and PI Impact.

The First Cheques Went to Schools, Hospitals and Trades

Wirz describes a three-link chain that used to be three products: use AI to learn a skill, use it to land a matching job, then use it to produce more in that job. “Today you can build a single layer, particularly within a vertical, that does all three,” he said. Fund III’s first three companies sit on that chain, not on a generic chatbot wrapper.

FUND III’S FIRST THREE BETS

Asked which of the three categories he would back with a last cheque, Wirz refused the split. “It’s like asking me to choose between my children,” he said. “The whole point is that these things are merging into each other, and a lot of the value is in the intersections.”

imagi Builds a Guardrail Between Models and Classrooms

Stockholm-founded, New York-based imagi raised a $4.5 million seed round on 23 July 2026, led by Brighteye and Day One Capital. The company has now raised $5.5 million. It does not train its own models. It sits between tools from OpenAI and Lovable and the school district, with screened prompts and a curriculum that teachers can actually run.

Co-founders Dora Palfi, the chief executive, and Beatrice Ionascu, the chief technology officer, say the platform already reaches more than 700,000 students in 140 countries, including more than 100 US districts and the Hawaii Department of Education, and more than 30,000 educators. Palfi called AI education “the new literacy” for children who will not know a world without it, and argued that the tools were not designed with children in mind.

That is the school end of HumanOS. The fund is buying the layer that lets a district deploy frontier tools without handing students a raw chat window.

NEX Flags Infection Risk Before It Spreads

NEX Health Intelligence is the hospital end. Brighteye’s Isabella Vahdati, now a principal, wrote that the firm is backing infection intelligence for hospital teams because most hospital tools still record outbreaks after they start. NEX reads lab results, ward transfers, contacts and prescribing in real time and ranks which patients are most at risk.

The model was built by co-founder Ashleigh Myall during a PhD at Imperial College London, then published in The Lancet Digital Health. Early clinical tests put the algorithm at an AUC of 0.89. Co-founder Dr Chang Ho Yoon is a physician and biomedical informatician. The product is live in two NHS Trusts in London, sites in North West England, and a military hospital in Southeast Asia, and has been used with more than 40,000 patients.

Healthcare-associated infections cost European health systems more than €24 billion a year, Vahdati wrote, and antimicrobial resistance is rising across more than 40 percent of the pathogen-antibiotic pairs the World Health Organization tracks. NEX is a productivity cheque dressed as clinical software: it sells time and judgement back to infection-control teams.

Gyver Matches Electricians the Way They Already Hire

Gyver, based in Brescia, is the labour-market end. Brighteye led its €1.4 million pre-seed on 13 May 2026, with āltitude, Vento Ventures, Zanichelli Venture, Antler and angels. Founders Francesco Defendi, Leo Acciarri and Mattia Zarrelli built a WhatsApp-first hiring platform for electricians, copying the referral traffic that already runs the trade.

The company has gathered a community of about 25,000 electricians. David Guérin, now a partner, said Europe has 2.7 million electrical workers and a €3 billion hiring market that has been “chronically underserved.” Gyver has also pointed to a need for 5.8 million extra skilled blue-collar workers in the EU by 2030 as grids, factories and buildings electrify. Early employer names circulating around the round include Schneider and Amazon.

Zanichelli Venture’s presence on that cap table matters for Fund III. The same publishing house is now a limited partner in the new vehicle. The electrician marketplace is not a side bet that wandered in from climate tech. It is the work half of a thesis the LP already touched.

European Learning Cash Doubled, Then Clustered at the Top

The firm’s own numbers make the raise look late, not early. Rhys Spence’s more than doubling from €710 million to €1.6 billion in European learning-and-work venture funding from 2024 to 2025 matched the highest yearly total since 2021. Deal count rose 21 percent to 334, the busiest year since 2018. Global funding in the same slice rose from €5.4 billion to €9.0 billion, with 1,400 deals against 1,403 in 2018.

EUROPEAN LEARNING AND WORK VC

H1 2026 already equals 87 percent of 2025’s full-year capital, on 53 percent of 2025’s deal count, and Brighteye says the run-rate is on pace to nearly double last year’s total. The catch inside those totals is concentration. Spence’s 2025 cut found the top three rounds, Amboss at €240 million, Lingokids at €103 million and Sdui at €95 million, summed to €438 million, about 61 percent of the top-10 total of €721 million. Sdui is a Brighteye holding. By H1 2026 the top 10 had risen to about 76 percent of funding, helped by Legora’s €525 million round.

The 2025 split under the firm’s own labels was €471 million into conventional edtech and €601 million into corporate and workplace learning. The biggest European cheques went to regulated markets and institutional buying, healthcare workflow, school infrastructure, staffing systems, not to nice-to-have engagement tools. The United Kingdom, Germany, France, Spain and the Netherlands accounted for over 80 percent of deals. Y Combinator was the most active learning-and-work investor globally in 2025.

That is a poor tape for a generalist writing $20 million growth tickets. It is a usable tape for a specialist writing $0.5 million to $4 million into the unfashionable plumbing of the same boom.

Who Backed the First Close

Wirz said the LP pitch was the portfolio’s own drift plus the track record. “We convinced them that our right to win comes from following what we’re already seeing in our portfolio.” The new money is not a Silicon Valley crossover syndicate. It is education capital and impact capital sitting down with the same public-market names that closed Fund II.

THE MONEY AROUND THE TABLE

• Lumina Foundation: Indianapolis education philanthropy whose venture arm already writes cheques into US edtech and social enterprises.
• Zanichelli: Italian education publisher; Zanichelli Venture had already joined Gyver’s pre-seed.
• PI Impact: the Rotterdam impact family office of the ProDelta group, with holdings in climate, materials and education access, now listed as a Fund III LP.
• Returning: the European Investment Fund, the Jacobs Foundation, and several European family offices.

Lumina’s seat is the clearest US education signal on the register. Zanichelli’s seat is the clearest publisher signal, and it already had a direct line into the electrician book. PI Impact’s seat fits a firm that now talks about labour volatility in energy, health and climate in the same breath as classrooms. None of those LPs needed a story about replacing teachers or clinicians. They needed a story about making those people more productive.

Most of the Money Will Sit in Cheques Under $4 Million

Fund I was €50 million. Fund II closed at €100 million in June 2023 and took then-AUM to €150 million. Fund III’s €100 million target is a flat sequel, not a step-up, even though the category’s headline funding has more than doubled. The extra room is in the number of names, up to 35, and in the earlier entry point.

THE FIRM’S PATH TO FUND III

1. 2017: Alex Spiro Latsis and Benoit Wirz found Brighteye, domiciled in Luxembourg, with research offices in London and Paris.
2. June 2023: Fund II closes at €100 million, double Fund I, with the EIF and the Jacobs Foundation on the register.
3. 13 May 2026: Brighteye leads Gyver’s €1.4 million pre-seed in Brescia.
4. 23 July 2026: Brighteye co-leads imagi’s $4.5 million seed for classroom AI literacy.
5. 16 September 2026: The firm announces a $72 million first close of Fund III and $245 million in AUM.

Existing labour marketplaces in the older funds, Zen Educate, Ornikar, Shakers and Installer, are the bridge. Fund III keeps writing into that pattern and into high-stakes software in healthcare, education, infrastructure, cybersecurity and climate. The World Health Organization’s own line, that healthcare-associated infections affect one in ten hospitalised patients, is the kind of ugly operational fact this fund wants on a term sheet.

Alongside the close, Brighteye promoted Guérin to partner, Vahdati to principal and Spence to head of platform and research. The team stays evenly split between investment and platform. An external bench of about 20 operator mentors sits around them. A new brand from Justified Studio arrived with the raise.

Uphill Taught the Firm to Sell Capacity Inside the Job

Wirz is blunt that the thesis followed the companies. He pointed to Uphill, a Fund II healthcare name that started as a learning platform and became a productivity layer inside hospitals. “What they realised was that people didn’t need a theoretical platform to learn. What they needed was knowledge in the flow of work,” he said. Hospitals were told, in his telling, “yes, we are training you, but actually what we’re selling you is capacity.”

That is the budget trick corporate learning never solved. Companies paid for courses and never trusted the return. A product that changes output on the ward, on the site, or in the classroom can be sold as labour, not as content. imagi’s district contracts, NEX’s infection scores and Gyver’s WhatsApp hiring all try to cash that cheque in different uniforms.

The defining opportunity of the AI era is not replacing humans with machines. It is expanding what humans can do with machines. Autonomous intelligence multiplies the potential value of human skills, judgement, and talent. The next generation of category-defining companies will use technology to redefine human capability. We call this the HumanOS and we’re partnering early with the founders building in this space.

Benoit Wirz, Founding Partner, Brighteye Ventures

His screen for founders is the problem, not the category label. “Are you trying to solve learning? Are you trying to solve employment? Are you trying to solve productivity? If you have a differentiated approach that is not using AI, we’re open to that.” The fund still has to finish raising. Final close is planned for the first half of 2027. Until then the $72 million is enough to keep writing the same kind of early, vertical cheques the first three names already represent.

The wager only pays if teachers, infection-control nurses and electricians can work with the models instead of being priced out by them. That is why a school safety layer and a Brescia hiring chat sit in the same vehicle, and why the firm did not ask LPs for a larger fund to make the point.

Disclaimer: This article is news reporting and analysis of a venture-capital first close and related portfolio activity. It is for information only and does not constitute investment advice, an offer, or a solicitation to commit capital to Brighteye Ventures, Fund III, or any company named above. Readers considering an investment in private funds or startup equity should consult a qualified financial adviser or regulated allocator who can review fund documents, fees and risk. Figures, close status, limited-partner lists and portfolio facts reflect the primary statements and data cited here and may change before the planned final close in the first half of 2027.