<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Educapital on Medium]]></title>
        <description><![CDATA[Stories by Educapital on Medium]]></description>
        <link>https://medium.com/@educapital_vc?source=rss-5c4f305f8e64------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*BdcCjt2eRioxazJoD3U_kA.png</url>
            <title>Stories by Educapital on Medium</title>
            <link>https://medium.com/@educapital_vc?source=rss-5c4f305f8e64------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 17:12:06 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@educapital_vc/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Edtech & Future of Work — September 2026]]></title>
            <link>https://medium.com/@educapital_vc/edtech-future-of-work-september-2026-2fad47517b8c?source=rss-5c4f305f8e64------2</link>
            <guid isPermaLink="false">https://medium.com/p/2fad47517b8c</guid>
            <dc:creator><![CDATA[Educapital]]></dc:creator>
            <pubDate>Tue, 13 Oct 2026 07:25:48 GMT</pubDate>
            <atom:updated>2026-10-13T07:25:48.958Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/311/1*2I6A4fBGCJC-Iwm5dIrzEg.png" /></figure><h3>Edtech &amp; Future of Work — September 2026 Snapshot</h3><ul><li>September 2026 saw a sharp spike in Edtech and Future of Work funding, with <strong>c.$1 143m raised globally, more than double the c.$503m raised in September 2025</strong>. The jump was driven by major rounds in <strong>Europe and the USA</strong>.</li><li>Notable raises include Amsterdam- and Tel Aviv-based <strong>Wonderful</strong> (<strong>$550M Series C</strong> led by Insight Partners at a $5B valuation), more than doubling the $2B valuation of its previous round six months earlier, which plans to accelerate product development and expand its forward-deployed engineering teams to meet demand for its enterprise AI operating system; <strong>HiBob</strong> (<strong>$166M</strong> led by Salesforce at a c.$3.2B valuation), which intends to use the capital partly for acquisitions and to turn its HR platform into an infrastructure layer for AI-powered workplaces; <strong>Jack &amp; Jill</strong> (<strong>$40M Series A</strong> led by Air Street Capital), which will use the funding to expand in the United States its hiring marketplace, where two AI agents represent candidates and employers; <strong>Loora</strong> (<strong>$22M Series B</strong> led by Union Tech Ventures), which plans to scale its AI research and engineering, accelerate go-to-market across the US, Europe, Asia and South America, and expand its research infrastructure; and Lyon-based <strong>Ed.ai</strong> (<strong>€5M Seed</strong> led by Bpifrance), whose AI assistant pre-corrects handwritten student work and builds personalised remediation paths, and which is preparing its expansion across the United States.</li><li><strong>Europe</strong> was the most active region this month, accounting for the majority of capital raised (55%), followed by <strong>North America</strong> (34%) and <strong>Asia</strong> (10%).</li></ul><h3>Some notable funding rounds around the world:</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/413/1*W315LnXBm2ZDodKtZKilow.png" /></figure><p><strong>M&amp;A Activity This Month</strong></p><ul><li>There were <strong>33 M&amp;A</strong> deals this month.</li><li>This month, <strong>Pearson</strong> announced it has agreed to acquire Workera, an AI-native skills intelligence platform that measures what employees can actually do through role-specific scenarios and simulations. Founded in 2019 in Palo Alto, <strong>Workera</strong> has around 60 employees and counts ServiceNow, Accenture and the United States Space Force among its customers. Expected to close in H2 2026, the deal will bring Workera into Pearson’s Enterprise Learning &amp; Skills unit.</li><li>In higher education, <strong>Tribal Group</strong> agreed a sale to <strong>Main Capital Partners</strong> for about £231.2 million, after the price was raised from £189.3 million in September. The deal is contested by Jenzabar, Tribal’s largest shareholder with 26.19% of the shares, which floated a possible rival offer at 111p per share. The shareholder vote is now scheduled for October 26.</li><li>Other notable deals included <strong>McGraw-Hill’s </strong>acquisition of <strong>TeachFX</strong>, an AI-powered instructional coaching platform, and <strong>Scholastic’s</strong> agreement to acquire children’s book publisher <strong>Cottage Door Press</strong>.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/343/1*0b4VpNiU_JyYjuWuvPUjFg.png" /></figure><p><strong>What’s new in AI — September 2026</strong></p><ul><li><a href="https://openai.com/index/introducing-dots/"><strong>OpenAI</strong></a> introduced Dots at its DevDay on September 29, new “always-on” AI agents powered by GPT-6 Astra, along with ChatGPT Space, a shared workspace where teammates and Dots agents work together.</li><li><a href="https://blog.google/innovation-and-ai/technology/ai/google-ai-updates-september-2026/"><strong>Google</strong></a> launched Gemini 4 Argon, its newest frontier model, capping a month of releases that also included Gemini 3.8 Flash.</li><li><a href="https://www.anthropic.com/claude-fable-and-mythos-5-1"><strong>Anthropic</strong></a> released Claude Fable 5.1 and Claude Mythos 5.1 on September 1, the same model with different levels of safeguards, setting a new standard for coding, knowledge work and long-running tasks.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jsFbhI3rWc8GmZviUiEyCQ.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2fad47517b8c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why we invested in Mallow]]></title>
            <link>https://medium.com/@educapital_vc/why-we-invested-in-mallow-54f29f3d64c2?source=rss-5c4f305f8e64------2</link>
            <guid isPermaLink="false">https://medium.com/p/54f29f3d64c2</guid>
            <dc:creator><![CDATA[Educapital]]></dc:creator>
            <pubDate>Mon, 12 Oct 2026 08:13:25 GMT</pubDate>
            <atom:updated>2026-10-12T08:13:25.238Z</atom:updated>
            <content:encoded><![CDATA[<p><em>Mallow has just launched its first toy: a screen-free, AI-powered audio games speaker that lets children aged 4 to 10 learn by talking, not scrolling. A couple of months ago, we backed Cédric and Flore to realise their vision and bring their unique product to market. Today, we are proud to see it reach families’ homes.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JZNLuOBwB5zGOr8AjASABg.png" /><figcaption><em>Cédric O and Flore Cousin, co-founders of Mallow. Photo: François Tancré / Les Echos.</em></figcaption></figure><p>Since 2018, we have had a strong conviction: technology can do a lot for children’s development:</p><ul><li><strong>For the youngest, without screens,</strong> by nurturing their creativity and language, as <a href="https://lunii.com/fr-fr">Lunii</a> does with its <em>Fabrique à Histoires</em>: a little box, with more than 2 million units sold, that lets children create their own stories with nothing but their imagination.</li><li><strong>As they grow older, through educational apps</strong> such as <a href="https://buddy.ai/fr/">Buddy.ai</a>, which helps them learn a new language through conversation.</li></ul><h3>Generative AI makes personalised interaction possible</h3><p>Every parent of a young child knows the story: screens are everywhere, and experts increasingly warn about their limits. The World Health Organization says no screen time before two. In France, the expert commission on children and screens went further in 2024: no screens before age 3, and only limited, accompanied use until age 6. At Educapital, our position has always been clear: no screens for the youngest, and a reasoned use of technology for the others.</p><p>But removing screens is not enough. Children learn language through back-and-forth exchanges, not passive listening. Academic studies show a strong difference in effect between the two. For instance, babies exposed to a foreign language through live interaction learn its sounds, while those hearing the same content from a recording learn nothing (<a href="https://www.pnas.org/doi/full/10.1073/pnas.1532872100">Kuhl et al., 2003</a>). Research from MIT and Harvard also shows that the number of conversational turns a child takes predicts activity in Broca’s area, the brain’s core language region (<a href="https://doi.org/10.1177/0956797617742725">Romeo et al., 2018</a>).</p><p>Until now, screen-free audio was a one-way experience: children were talked to, rarely talked with. Generative AI and voice technology change that. For the first time, audio can listen, respond and adapt to each child.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*po1CpLx7EFYcKd44Fj6HxQ.png" /><figcaption><em>Photo: Mallow.</em></figcaption></figure><h3>Mallow, the next generation of edutainment hardware</h3><p>Mallow is creating a new way to play and learn: audio games. In video games, children play behind a screen, controller in hand. With Mallow, there is no screen: their voice becomes the controller. A child places a figurine on the speaker, and the adventure begins. They talk, make choices, solve problems and invent, becoming the hero of their own story. They embark on missions that carry an educational aim, such as understanding emotions.</p><p>Like Lunii, Mallow nurtures children’s creativity without a screen. This time, it integrates AI to personalise the experience to each child and get them to interact, which is what drives language development.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZDy04gQqMHB-AkvH0ig2fg.png" /><figcaption><em>The Mallow speaker with its Miraculous figurine. Photo: Mallow.</em></figcaption></figure><p>Mallow was built hand in hand with families. Over two years, 35 creatives, engineers, researchers and business profiles tested Mallow with more than 300 families, across 6,000 play sessions. In a category where trust is everything, that rigour makes all the difference.</p><p>And the result is a product with unique characteristics that perfectly meet a child’s needs:</p><ul><li><strong>AI stays in the background.</strong> Stories, rules and characters are written by children’s authors, narrative designers and cognitive science experts, with a scientific council of psychology and child-development specialists.</li><li><strong>Models built for children’s voices.</strong> Several AI models, including Mistral AI, understand hesitations, lisps, even a child talking with a pacifier, and adapt to each child’s language level.</li><li><strong>A game device, not a full companion.</strong> Every story has a beginning, an end and objectives. Characters remind children they are imaginary, and anything worrying stops the game and points the child to a trusted adult.</li><li><strong>Safety first. </strong>Mallow never asks for a child’s address, school or contact details, and the team constantly tests its own guardrails through simulations, never with real children. It also applies the AI Act and the Cyber Resilience Act from day one, ahead of their deadlines.</li></ul><h3>Born from an exceptional and complementary founding team</h3><p>We backed Mallow one month after its creation, and Cédric and Flore were the first reason why. Together, they bring exactly what “inventing a new category of product”, in Flore’s words, requires.</p><p><a href="https://www.linkedin.com/in/cedric-o/">Cédric O</a> needs no introduction. Having shaped France’s digital policy and co-founded Mistral AI, he knows better than anyone what is at stake when AI meets children. For Mallow, he surrounded himself with two top-tier profiles:</p><ul><li><a href="https://www.linkedin.com/in/flore-cousin-09449631/">Flore Cousin</a>, co-founder and COO, is his perfect complement: where Cédric brings the vision and deep AI expertise, she brings the operational rigour, sharpened at BCG, P&amp;G and Alma, that it takes to turn a breakthrough product into one parents trust. With seven children under ten between Cédric and Flore, the least we can say is that they know their persona first-hand!</li><li><a href="https://www.linkedin.com/in/boyerseb">Sébastien Boyer</a>, non-executive co-founder, brings deep expertise in AI-powered hardware, from prototype to production. He previously founded FarmWise, an AI and robotics company that raised hundreds of millions of dollars and was acquired in 2024. He is also an independent board member of <a href="https://powerz.tech/fr-fr/">Powerz</a>, an edutainment company in our portfolio.</li></ul><p>Their vision attracted top-tier backing: within weeks of its founding, Mallow raised €11M in a Seed round led by General Catalyst and Balderton Capital.</p><h3>What’s next?</h3><p>Our dream is simple: a generation of children who talk, imagine and invent more, and scroll less. With Mallow, it starts this Christmas: the team hopes to find its way under hundreds of trees, with new figurines and stories to follow. A bilingual version is in the works, and Mallow hopes to launch in the US as early as next year.</p><p>For us, Mallow is more than a new investment: it closes a loop. As Litzie Maarek, co-founder of Educapital, puts it:</p><blockquote>“Eight years ago, Lunii showed us that parents want screen-free experiences for their children. Mallow brings those experiences to life through conversation: children don’t just listen anymore, they talk, imagine and build their language.”</blockquote><p>We are proud to support Cédric, Flore and the whole Mallow team on this journey. This is only the beginning.</p><p><em>NB: For those interested, Mallow is available for €119.99 at </em><a href="https://www.mallow.fr/"><em>mallow.fr</em></a><em>.</em></p><h4><strong>About Educapital</strong></h4><p><em>Educapital is one of Europe’s largest impact VC funds dedicated to education and the future of work. With €200M under management and around 40 portfolio companies, we invest €1–7M tickets from Seed to Series B in startups that help people learn, work and grow.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=54f29f3d64c2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Edtech & Future of Work — June 2026]]></title>
            <link>https://medium.com/@educapital_vc/edtech-future-of-work-june-2026-30185cda1eed?source=rss-5c4f305f8e64------2</link>
            <guid isPermaLink="false">https://medium.com/p/30185cda1eed</guid>
            <dc:creator><![CDATA[Educapital]]></dc:creator>
            <pubDate>Thu, 09 Jul 2026 12:53:32 GMT</pubDate>
            <atom:updated>2026-07-09T12:53:32.848Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/295/1*U8InS8MT-FYjaGHobIFScQ.png" /></figure><h3>Edtech &amp; Future of Work — June 2026 Snapshot</h3><ul><li>June 2026 saw a sharp spike in Edtech and Future of Work funding, with <strong>c.$768m raised globally, up from c.$637m in June 2025</strong>. The jump was driven by major rounds in the<strong> USA and Spain.</strong></li><li>Notable raises include Barcelona-based <strong>Factorial </strong>(<strong>$150M</strong> <strong>Series D</strong> led by General Catalyst at a $2.5B valuation), which plans to expand across Europe, with a new Munich office anchoring its push into Germany, as it shifts from HR SaaS to an AI-first workforce operations platform; <strong>Warp ($60M Series B)</strong>, which intends to build out its agentic coding platform that deploys AI agents to automate software engineering tasks; <strong>Town ($55M Series A)</strong>, which plans to scale its personal AI assistant that proactively manages email, calendar, and daily workflows; <strong>Stepful ($55M Series C)</strong>, which will use the funding to scale its AI-powered platform training and placing entry-level workers into healthcare jobs; and<strong> Convey ($38M Series A</strong> led by Andreessen Horowitz), which plans to build out its AI teammate platform automating back-office workflows for enterprise operators.</li><li><strong>North America</strong> was the most active region this month, accounting for the majority of deals and capital raised (<strong>52%</strong>), followed by <strong>Europe</strong> (<strong>31%</strong>) and <strong>Asia</strong> (<strong>18%</strong>).</li></ul><h3>Some notable funding rounds around the world:</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/394/1*5NHHegUsYL61FtBUwUaPzQ.png" /></figure><p><strong>M&amp;A Activity This Month</strong></p><ul><li>There were <strong>32 M&amp;A deals</strong> this month.</li><li>This month, <strong>Ornikar</strong> announced the acquisition of <strong>En Voiture Simone</strong>, a France-based online driving-school platform, creating a group with over 10% of the French driving-school market and combined revenue nearing €150M.</li><li><strong>Superhuman</strong> (formerly Grammarly) also announced the acquisition of <strong>GPTZero</strong>, a US-based AI-content detection platform with 19M+ users and $30M ARR, to build out an “authenticity layer” across its productivity suite.</li><li>Other notable deals included <strong>Superprof</strong>’s acquisition of <strong>Profes</strong>, a Brazil-based tutoring marketplace, as part of its international roll-up strategy, and <strong>mthree</strong>’s acquisition of <strong>CAPSLOCK</strong>, a UK-based cybersecurity training provider, to strengthen its reskilling and Hire Train Deploy offering.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/305/1*kKxhYS3EUIOeQ8tcvuG7DQ.png" /></figure><p><strong>What’s new in AI — June 2026</strong></p><ul><li><a href="https://www.anthropic.com/news/confidential-draft-s1-sec"><strong>Anthropic</strong></a> and <a href="https://www.theglobeandmail.com/investing/markets/indices/TXLV/pressreleases/3116489/openai-anthropic-ipo-what-s-confirmed-what-s-still-speculation/"><strong>OpenAI</strong></a> both <strong>confidentially filed for IPOs</strong> in June, with Anthropic moving first on June 1 at a reported $965B valuation, setting up a historic race between the two leading AI labs to reach public markets later this year.</li><li><a href="https://www.anthropic.com/news/claude-fable-5-mythos-5"><strong>Anthropic</strong></a> launched <strong>Claude Fable 5 and Claude Mythos 5</strong>, its new top-tier “Mythos-class” models sitting above Opus with state-of-the-art performance in coding and long-horizon agentic work, before briefly suspending both on June 12 to comply with a US export-control order, and restoring global access to Fable 5 on July 1 once the controls were lifted.</li><li><a href="https://www.marktechpost.com/2026/06/18/perplexity-launches-brain/"><strong>Perplexity</strong></a> launched <strong>Brain</strong>, a self-improving memory system for its Computer agent that builds a context graph of the agent’s past work and reviews it overnight to improve accuracy and cost on future tasks.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*r7LRViOmDxTmMVeNSckbwA.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=30185cda1eed" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why we Invested in Rocapine]]></title>
            <link>https://medium.com/@educapital_vc/why-we-invested-in-rocapine-b6c3465b5787?source=rss-5c4f305f8e64------2</link>
            <guid isPermaLink="false">https://medium.com/p/b6c3465b5787</guid>
            <dc:creator><![CDATA[Educapital]]></dc:creator>
            <pubDate>Wed, 08 Jul 2026 12:41:11 GMT</pubDate>
            <atom:updated>2026-07-08T12:42:24.448Z</atom:updated>
            <content:encoded><![CDATA[<p>By <a href="https://medium.com/u/8aed6a0cd7e4">Alexandre Glaser</a> and <a href="https://medium.com/u/78c11f40b6bc">Alexis Pavy</a></p><p><em>Educapital is proud to lead Rocapine $13M Series A, along other investor such as Ring Capital.</em></p><p>Structured wellness and personal development is still, for most people, a luxury they can’t afford. The professionals who actually help us learn, live better and grow, coaches, therapists, personal trainers, nutritionists, are delivered one human at a time, and priced accordingly: roughly $75–250 an hour in the US, €40–150 in Europe. Sustained self-improvement therefore behaves like a premium good, reserved for those with the money, the time and the network to reach it. The result is a self-improvement gap that mirrors the wealth gap: the people who would benefit most from a coach or a mentor are precisely the ones who can least afford one.</p><p>Meanwhile, we spend years of our attention on our phones, on mediocre content engineered to keep us scrolling, and almost none of it compounds into anything. The average adult now spends around 4 hours 37 minutes a day on their phone, closer to six hours for Gen Z. Over a lifetime that adds up to somewhere between 8 and 20-plus years. It is the single largest block of discretionary time most of us will ever have, and the overwhelming majority of it evaporates the moment we lock the screen.</p><p>Enter <strong>Rocapine</strong>: an AI-first wellness app studio and publisher on a mission to redirect that time, away from passive, anxiety-inducing consumption and toward applications that make users feel genuinely more alive.</p><h3>A Deep, Fast-Growing Market</h3><p><strong>Mobile application is a structurally large and durable category</strong>, everyone knows that. The more interesting point is where attention and money are moving <em>within</em> it. The centre of gravity is shifting away from pure entertainment: content-based, non-gaming apps are steadily gaining share as users spend more of their time and their wallet on wellness, productivity and self-improvement rather than on games. This is a durable behavioural shift, not a fad, and it sits on top of the largest block of discretionary time most people have.</p><p>The depth is real. The broader market for premium, subscription-driven non-gaming apps represents <strong>$85Bn in spending in 2025</strong>. A figure that is expected to grow to <strong>$100b</strong>n <strong>in 2030</strong>, and will keep growing as consumers pay for value-adding digital services. Within that, the wellness segment has itself expanded into a <strong>$27bn-plus market</strong>, carried by strong structural tailwinds.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*HgUZIM-2l8ARxOEkDuM--g.png" /><figcaption>Sensor Tower State of Mobile Report- Annual Spending for Non Gaming and Gaming Mobile Application</figcaption></figure><p><strong>Two dynamics compound here</strong>, the rise of premium non-gaming experiences and the rapid expansion of digital wellness, and together they open up a large, diversified space. That is exactly the kind of deep, growing pool in which a hit-driven publisher can build something very large, launching many apps that each compete in their own sub-market while sharing one engine underneath.</p><h3>Ready to be Disrupted by a New Technology</h3><p>Over the last decade we watched several players, Voodoo for example, industrialise the creation of mobile apps: streamlining and scaling production on the back of hard data proof-points.</p><p><strong>Yet that playbook never really crossed over into wellness</strong>. The reason is simple. Wellness apps are content-heavy, and until recently content required two expensive resources: time and a dedicated team. Creating, testing and scaling could only happen at a slower pace than in other verticals, so the hit-factory model that worked in gaming never took hold here.</p><p>Generative AI removes that bottleneck. With the right AI stack you can produce content at a pace never seen before, test features and onboarding flows, and iterate at genuinely high velocity. This is the core of Rocapine’s edge. Its proprietary stack, <strong>RocaStudio</strong> for the build cycle, <strong>RocAI</strong> for content and intelligence, and <strong>RocaBoost</strong> for marketing and monetisation, automates the key steps across product development, user acquisition and monetisation, giving an independent studio the same velocity as a major one.</p><p>What convinced us the model is real, not theoretical, is how much it compresses the build. Rocapine has taken time-to-MVP from more than three months down to under two weeks, and dramatically shortened the time it takes a new app to become a hit:<strong> from an industry norm measured in years to a matter of weeks.</strong></p><p>It operates as both an <em>editor</em>, identifying and co-developing high-potential concepts, and a <em>publisher</em>, industrialising launch and scale through its platform. Effectively, Rocapine is building the “Voodoo of wellness”: a repeatable, AI-native engine for launching and scaling wellness apps, whether developed in-house or brought in by independent studios it equips with its tools.</p><h3>Built by a Stellar Team: When Clarity of Vision meets Velocity of Execution</h3><p>During our first exchange with the team, we were impressed by their knowledge of the market, but most and foremost the clarity of their vision. Stan, Jean-Gabriel and Sammy have an impressive ability to articulate the issue they want to solve and the disruption they are bringing to the market, both from a technological and operational standpoint.</p><p><strong>And they have the numbers to back up their claim</strong>. In less than a year, they tested more than 100 concepts, and already built an impressive portfolio of wellness applications in a wide area of segments, Harmony, Eve, Unchaind, Stashcook… Already showcasing operationaly and from and end-user point-of-view (with an average rating on the stores of +4,5/5 for their portfolio) the value of their system.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*aEg6kb7YOPmjEu4Peeizxg.jpeg" /></figure><p><strong><em>About Educapital</em></strong><br><em>Educapital is the largest European impact investment fund dedicated to the Future of Education and Work. Founded in 2017 by Marie-Christine Levet and Litzie Maarek, Educapital invests from late seed to Series B in European start-ups that are disrupting or innovating in these sectors. Educapital invests in every field of Edtech — early childhood, K12, higher education, continuing education — , and focuses on HRtech solutions dedicated to talent first — professional training, recruitment and mobility, engagement and well-being in the workplace.</em><br><em>Educapital’s ambition is to help European champions emerge by providing them with initial financial support, ongoing expertise and market-specific know-how.</em><br><em>Educapital has invested in more than 45 leading companies, including Labster, 360 Learning, Preply, Lunii and Knowunity, making it one of the most active Edtech and Future of Work funds in the world.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*REWwPyppvjsYJ-0i.png" /></figure><p><em>This operation benefits from support from the European Union under the InvestEU Fund.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b6c3465b5787" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why we Invested in Mendo]]></title>
            <link>https://medium.com/@educapital_vc/why-we-invested-in-mendo-11643f7eadd1?source=rss-5c4f305f8e64------2</link>
            <guid isPermaLink="false">https://medium.com/p/11643f7eadd1</guid>
            <dc:creator><![CDATA[Educapital]]></dc:creator>
            <pubDate>Wed, 24 Jun 2026 09:27:56 GMT</pubDate>
            <atom:updated>2026-06-24T09:27:56.754Z</atom:updated>
            <content:encoded><![CDATA[<p>By <a href="https://medium.com/u/2e9b0d69d091">Litzie Maarek</a>, <a href="https://medium.com/u/8aed6a0cd7e4">Alexandre Glaser</a> and <a href="https://medium.com/u/c521d1ffc9e6">Alexis Pavy</a></p><p><em>Educapital is proud to co-lead Mendo €12M Series A, along other investor such as Ventech Capital.</em></p><p>Enterprises are pouring billions into generative AI, yet most of that spend is quietly evaporating. The problem isn’t the technology in itself, it’s adoption. 86% of employees want training in generative AI, but only 14% have received any (<a href="https://www.shrm.org/topics-tools/news/hr-magazine/ai-employee-training#:~:text=A%202023%20international%20survey%20by,job%20postings%20requiring%20AI%20skills">source</a>), and 64% of knowledge workers report little to no familiarity with the GenAI tools meant for their roles (<a href="https://asana.com/fr/resources/state-of-ai-work-2024">source</a>). Companies are buying Copilot and agent licenses at scale and have almost no idea whether anyone is using them, or whether a euro of ROI has come back.</p><p>This gap is the defining enterprise software problem of the moment, and it is getting wider as the agent wave accelerates. A company can roll out Claude or Copilot organization-wide overnight, getting twenty thousand employees to actually use them well is a transformation problem, not a procurement one. That is the problem Mendo solves.</p><h3><strong>A training layer that lives in the flow of work</strong></h3><p>Most employees aren’t natural early adopters, and they don’t have time to learn a new tool on the side. Mendo doesn’t ask them to. Rather than sending people to a separate learning platform, Mendo embeds directly into the tools they already use, their browser-based AI chats and their Microsoft 365 environment, and guides them step by step, from a first basic prompt to building agents for specific use cases. It meets users where they already work, which is the only place adoption actually happens.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Tn3Kb2Vd7BO_GkW7_aCOOg.png" /></figure><h3><strong>An analytics layer that proves it worked</strong></h3><p>Once employees are using AI and building agents, the questions start: Which agents are actually used? Which deliver real value? Did we save money? These are the questions a CFO asks of any eight-figure investment, and today most companies can’t answer them. Mendo’s analytics layer gives managers and AI leads a comprehensive, cross-tool view of agent adoption, usage intensity, and measurable ROI, mapping the entire AI landscape deployed inside the company.</p><p>This is the part we find most compelling. The training layer is the wedge that gets Mendo deployed to every employee, the resulting usage data is the durable asset. It produces the one dashboard the C-suite can’t get from Microsoft or OpenAI, because Mendo sits across all of them, including internal GPTs, and it grows more valuable the more the company uses it.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*iguXmIS5dFGqhpyzKMntbA.png" /></figure><h3><strong>The perfect team for the project</strong></h3><p>We first met Quentin and Alexandre in 2021, and we’ve watched them up close ever since. They started with a single training module for Excel and have since built Mendo into the enterprise GenAI adoption platform it is today, expanding the product layer by layer with a speed of execution and an instinct for where the market was heading that we rarely see.</p><p>The proof is in who trusts them: Mendo already works with EY, PwC, Novo Nordisk, SNCF, Engie, Crédit Agricole, La Poste and Groupe SEB, among more than 100 enterprise clients. Earning that roster at this stage is rare, and it is the clearest signal of the obsession and execution that made us want to back them.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*L7wuk0ZBwdhHHNHstsvYIg.jpeg" /></figure><p><strong><em>About Educapital</em></strong><br><em> Educapital is the largest European impact investment fund dedicated to the Future of Education and Work. Founded in 2017 by Marie-Christine Levet and Litzie Maarek, Educapital invests from late seed to Series B in European start-ups that are disrupting or innovating in these sectors. Educapital invests in every field of Edtech — early childhood, K12, higher education, continuing education — , and focuses on HRtech solutions dedicated to talent first — professional training, recruitment and mobility, engagement and well-being in the workplace.</em><br><em> Educapital’s ambition is to help European champions emerge by providing them with initial financial support, ongoing expertise and market-specific know-how.</em><br><em> Educapital has invested in more than 45 leading companies, including Labster, 360 Learning, Preply, Lunii and Knowunity, making it one of the most active Edtech and Future of Work funds in the world.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wHPn4FXhQVE2zEGdAHmGSg.png" /></figure><p><em>This operation benefits from support from the European Union under the InvestEU Fund.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=11643f7eadd1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Edtech & Future of Work — April May 2026]]></title>
            <link>https://medium.com/@educapital_vc/edtech-future-of-work-april-may-2026-c1c283b2464f?source=rss-5c4f305f8e64------2</link>
            <guid isPermaLink="false">https://medium.com/p/c1c283b2464f</guid>
            <dc:creator><![CDATA[Educapital]]></dc:creator>
            <pubDate>Thu, 11 Jun 2026 08:05:11 GMT</pubDate>
            <atom:updated>2026-06-11T08:05:11.930Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/656/1*7EjAByyD1A-px3dHTIrP1w.png" /></figure><h3>Edtech &amp; Future of Work — April and May 2026 Snapshot</h3><ul><li>April and May saw a sharp spike in Edtech and Future of Work funding, with <strong>$1 711m</strong> raised globally, up from <strong>$1 260m in the same period last year. </strong>The jump was driven by major rounds in the USA.</li><li>Included San Francisco-based Factory ($150M Series C led by Khosla Ventures) plans to accelerate research into autonomous software engineering and expand its AI coding Droids platform globally, with new offices in London and Tokyo, and Hightouch ($150M Series D at a $2.75B valuation, co-led by Goldman Sachs and Bain Capital Ventures) intends to build out its agentic marketing platform combining a customer data layer, brand context, and cross-channel orchestration so AI agents can autonomously identify opportunities, generate on-brand creative, and execute campaigns at scale.</li><li>North America was the most active region this month, accounting for the majority of the deals and capital raised (64%).</li><li>The United Kingdom and Germany based startups <strong>Multiverse and Synera</strong> accounted for the largest rounds in Europe, respectively <strong>$70m</strong> and <strong>$40m</strong>.</li></ul><h3>Some notable funding rounds around the world:</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/902/1*PF8e-qJueJKUeSNgmH-lXw.png" /></figure><h3>M&amp;A Activity This Month</h3><ul><li><strong>49 exit deals</strong> were recorded in EdTech and the Future of Work in April and May 2026.</li><li>During this period,<strong> OpenAI acquired </strong><a href="https://techjacksolutions.com/ai-brief/openai-acquires-edtech-platform-chalkie-ai-in-second-vertica/"><strong>Chalkie AI</strong></a><strong> </strong>for an undisclosed amount, an AI-powered lesson-planning platform serving ~10m students and 500k teachers globally, signaling the company’s growing interest in the education vertical.<strong> </strong><a href="https://www.execonline.com/news/execonline-acquires-teamraderie/"><strong>ExecOnline</strong></a><strong> acquired </strong><a href="https://www.execonline.com/news/execonline-acquires-teamraderie/"><strong>Teamraderie</strong></a> for an undisclosed amount, a provider of expert-led team development experiences focused on AI transformation, to bridge individual leadership learning with measurable team-level outcomes.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/690/1*b92ghYKAuo8sqM5biMFVpA.png" /></figure><p><strong>What’s new in AI — April and May 2026</strong></p><ul><li><strong>OpenAI </strong>released GPT-5.5, the first fully retrained base model since GPT-4.5, posting a 60% reduction in hallucinations over its predecessor and claiming the #1 spot on Terminal-Bench 2.0 at 82.7%.</li><li><strong>Anthropic</strong> launched Claude Opus 4.7, recording the biggest single-version improvement on SWE-bench Pro of any model in 2026, jumping from 53.4% to 64.3% and reclaiming the coding crown.</li><li><strong>DeepSeek</strong> released V4 Preview, a 1.6 trillion parameter open-source model built on zero Nvidia hardware, priced at $0.14 per million input tokens, accelerating the pricing collapse in frontier AI.</li><li><strong>Google </strong>overhauled its Search engine in May, powering the search bar entirely with Gemini 3.5 Flash to generate AI-summarised pages rather than traditional link lists, marking the biggest change to Search in over 25 years.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c1c283b2464f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Payroll in the Age of AI]]></title>
            <link>https://medium.com/e-d-volution/payroll-in-the-age-of-ai-2f675a49e1f3?source=rss-5c4f305f8e64------2</link>
            <guid isPermaLink="false">https://medium.com/p/2f675a49e1f3</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[payroll]]></category>
            <category><![CDATA[hrtech]]></category>
            <category><![CDATA[vc]]></category>
            <dc:creator><![CDATA[Educapital]]></dc:creator>
            <pubDate>Tue, 14 Apr 2026 11:55:36 GMT</pubDate>
            <atom:updated>2026-04-14T12:19:05.812Z</atom:updated>
            <content:encoded><![CDATA[<p><em>By </em><a href="https://medium.com/u/26dc259ab795"><em>Géraldine Fillet</em></a></p><p><em>Payroll has always been one of those unglamorous, mission-critical processes that nobody wanted to touch. Regulatory complexity, zero-tolerance on errors, monthly batch cycles running like clockwork for decades, the incentive to disrupt was low, and the risk of breaking something was high. The launch of mass market LLMs in late 2022 opened a breach.</em></p><h4>The third wave of SaaS disruption</h4><p>The first generation of payroll SaaS (<a href="https://www.fr.adp.com/">ADP</a>, <a href="https://www.sage.com/fr-fr/paie-logiciels-gestion/">Sage</a>,..) essentially digitized the spreadsheet: same calculation logic, but in a web interface. Useful, but not transformative.</p><p>Then came the natural language layer : <a href="https://www.fr.adp.com/">ADP</a> launched <em>Roll by ADP</em> in 2023, letting managers run payroll by mobile conversation. <a href="https://www.workday.com/">Workday</a> and <a href="https://www.dayforce.com/cf">Ceridian</a> followed with AI copilots on their modules. Smarter to interact with, but the underlying engine remained the same rigid 1994 architecture, running once a month, unable to adapt in real time.</p><p>What’s happening now is a more fundamental shift. A new generation of players (<a href="https://www.warp.co/">Warp</a>, <a href="https://centralhq.com/">Central</a>, <a href="https://www.every.io/">Every.io</a>) is rebuilding the entire stack around autonomous agents. Their key differences?</p><ul><li><strong>Reactive vs. proactive</strong>: old tools waited for a human instruction, agents detect an event (new hire, change of tax residence, legislative modification) and act without being asked.</li><li><strong>Batch vs. real-time</strong>: agentic payroll runs continuously, not once a month. This enables immediate error detection and permanent commission visibility.</li><li><strong>Coded rules vs. reasoning</strong>: a rules engine breaks on exceptions (an employee worked across 3 states this month). An LLM agent reasons about the case, consults applicable regulations, and calculates correctly.</li><li><strong>Native explainability</strong>: new architectures automatically produce a narrative audit trail of each calculation, exactly what the EU 2027 directive will require (<a href="https://www.consilium.europa.eu/en/policies/pay-transparency/">learn more</a>).</li></ul><h4>Incumbents are not standing still</h4><p><a href="https://www.fr.adp.com/">ADP</a> serves more than one million companies and processes payroll for 40 million employees in the US. <a href="https://www.workday.com/">Workday</a>, <a href="https://www.dayforce.com/">Dayforce</a> (ex Ceridian) and <a href="https://www.sap.com/france/products/hcm.html">SAP HCM</a> manage payroll systems for the 500 largest companies worldwide.<br>These players are not blind to the transformation, they are investing massively (<a href="https://www.adp.com/what-we-offer/products/lyric.aspx">ADP Lyric</a>, <a href="https://newsroom.workday.com/2024-09-17-Announcing-Workday-Illuminate-TM-The-Next-Generation-of-Workday-AI">Workday Illuminate</a>,...). But their constraint is structural:</p><ul><li>M<strong>igrating</strong> one million clients to an agentic architecture in 18 months without a catastrophic risk of failure is impossible.</li><li><strong>The partner networks</strong> that incumbents rely on, accounting firms and integrators, are organized around the old batch workflows, and agentic payroll would disintermediate them.</li><li><strong>Their economic model</strong> often relies on the friction (human support) that agentic eliminates.</li></ul><p>This innovator’s dilemma opens a window for a new generation of players. Let’s dive into the major trends that we have seen in this market.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6AqB3Eb9hf8K9o1SJzCGfA.png" /><figcaption>Educapital’s proprietary landscape</figcaption></figure><h4>#1 — Global Payroll &amp; Employer of Record (EoR)</h4><p><em>How to legally pay and manage employees in multiple countries without creating a local legal entity?</em></p><p>These platforms combine multi-country payroll, local social and fiscal compliance, and, for EoR, legally employ the worker on behalf of the client company. They address structural demand driven by the rise of remote work and the globalization of teams.</p><p>Before 2019, international payroll management for non-multinational companies hit a wall: either create a legal entity in each target country (6 months, €20 000 to €50 000 per country), or rely on <a href="https://www.adp.com/what-we-offer/products/adp-globalview-payroll.aspx">ADP GlobalView</a>, <a href="https://www.oracle.com/fr/human-capital-management/">Oracle HCM Cloud</a> or<a href="https://www.sap.com/france/products/hcm.html"> SAP SuccessFactors</a>, systems designed for large enterprises, with 12–18 month implementation cycles. SMEs and startups were simply excluded from the market.</p><p><a href="https://www.deel.com/">Deel</a> and <a href="https://remote.com/fr-fr/">Remote</a> broke this model in the early 2020s with three simultaneous disruptions:</p><ul><li><strong>A transparent, self-serve pricing</strong> (Deel displays $599/month per EOR employee, accessible in a few clicks vs. ADP’s mandatory sales call).</li><li><strong>Proprietary entities</strong> in 80–150 countries vs. subcontracted partner networks</li><li>And an <strong>unmatched onboarding speed </strong>(<a href="https://www.deepl.com/fr">Deel</a> claims international onboarding in 2 minutes vs. 9 months for SAP).</li></ul><p>The result? Deel reached over $1 billion ARR in under 5 years, with more than 35 000 clients (primarily 10 to 500-employee companies that legacy systems simply didn’t serve)*. They didn’t steal the incumbents’ market, they created a new one. This market opportunity is now well funded with <a href="https://www.deepl.com/fr">Deel</a> and <a href="https://www.rippling.com/">Rippling</a> having raised over $750M in the last 18 months*.</p><p>Competition now plays out on the geographic breadth of proprietary entities and depth of product integration.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wxTeguuVyvywFmT5xaXkpw.png" /><figcaption>Overview of Deel’s platform</figcaption></figure><h4>#2 — Incentive Compensation Management (ICM)</h4><p><em>How to calculate, visualize and optimize in real time the variable compensation of sales teams?</em></p><p>This segment addresses specifically the RevOps, Finance and Sales Ops teams that control commission plans in commercial organizations. The ICM market is estimated at $10.25 billion in 2024, with $2,000–3,000 billion in commissions distributed annually worldwide. US players are dominating (<a href="https://www.xactlycorp.com/">Xactly</a>, <a href="https://www.captivateiq.com/">CaptivateIQ</a>, <a href="https://www.quotapath.com/">QuotaPath</a>, <a href="https://spiff.com/">SPIFF</a>), while European one are emerging (<a href="https://www.qobra.co/">Qobra</a>, <a href="https://www.amalia.io/">Amalia.io</a>, <a href="https://www.heydolfin.com/">Dolfin</a>, <a href="https://www.remuner.com/">Remuner</a>, <a href="https://www.palettehq.com/">Palette</a>).</p><p><a href="https://www.salesforce.com/news/stories/salesforce-signs-definitive-agreement-to-acquire-spiff/">The Spiff acquisition by Salesforce</a> (for $419M, in 2024) anchors the thesis that ICM will be absorbed by major suites (<a href="https://www.sap.com/france/index.html">SAP</a>, <a href="https://www.workday.com/fr-fr/homepage.html">Workday</a>, <a href="https://www.salesforce.com/fr/">Salesforce</a>). Pure-plays will need to either reach critical mass or specialize on a segment the giants will never serve well, notably European mid-market.</p><h4><strong>#3 — Pay Transparency</strong></h4><p><em>How to explain in real time salary gaps within a company in an ojective way?</em></p><p><a href="https://www.consilium.europa.eu/en/policies/pay-transparency/">The European 2027 Directive</a> and the <a href="https://disa.com/news/pay-transparency-us-legislation-updated-2025/">American legislative wave</a> are creating an entirely new software segment from scratch: pay transparency management. Companies with 100+ employees will need to publish compensation gaps by job category, justify them, and respond to individual requests.</p><p>Without a dedicated tool, this is impossible. Indeed, existing HR spreadsheets don’t have the required granularity. In this market, the risk is that major HCM editors absorb this functionality as standard. But this is also an opportunity for an efficient path to liquidity, for the leading pure players (<a href="https://ravio.com/">Ravio</a>, <a href="https://synd.io/">Syndio</a>, <a href="https://www.beqom.com/fr-fr">Beqom</a>, <a href="https://www.payanalytics.com/">PayAnalytics</a>).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*RIfz4QEjfh5GQHcqMP2GTQ.png" /><figcaption>Overview of Syndio’s tool for compensation</figcaption></figure><p>As a consequence, the global compensation analytics market is estimated at $3.7 billion in 2025, with an 8% CAGR through 2030*.</p><h4>#4 — Earned Wage Access (EWA)</h4><p><em>How to allow employees to access their already-earned wages before payday, without destabilizing the employer’s cash flow?</em></p><p>EWA is a quiet revolution in the relationship to pay: rather than waiting for end of month, employees, often hourly or precarious workers, can immediately unlock the share of their salary already earned.</p><p>The global EWA market is now valued at $30 billion, with a projected 23% CAGR through 2033*! Also the sector already benefit from a major regulatory green light, when, in the US, the CFPB (Consumer Financial Protection Bureau) clarified in December 2025 that EWA models backed by actual payroll are not “credit”.</p><p>Unsurprisingly, some US &amp; UK actors are already well funded (<a href="https://www.rainapp.com/">Rain</a>, <a href="https://stream.co/en/solutions/earned-wage-access">Stream</a>, <a href="https://kashable.com/">Kashable</a>), as those markets are more mature when it comes to benefits. Other western Europeans are developping well (see <a href="https://payflowfrance.webflow.io/">Payflow</a>).</p><p>What’s interesting to see is that the EWA category is evolving toward “integrated financial wellness.” After instant salary draw, come payroll-backed loans, employee credit cards, and AI coaching.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*WIPZvj0ymLT922YK4cqVYQ.png" /><figcaption>Overview of Payflow’s interface</figcaption></figure><h4>#5 — Payroll-as-a-service</h4><p><em>How to enable any software editor (vertical SaaS, ERP, accounting tool) to offer payroll to its own clients, without building it from scratch?</em></p><p>Embedded payroll is the infrastructure-as-a-service version of payroll: headless APIs that editors consume to embed payroll functionality into their own interfaces. The model follows the trajectory of <a href="https://stripe.com/">Stripe</a> in payments or <a href="https://plaid.com/open-banking/">Plaid</a> in open banking.</p><p>The payroll API market is estimated at $3.2 billion in 2026, with an 11.7% CAGR through 2030*. The underlying investment thesis is “payroll-as-infrastructure”. Every vertical SaaS managing workers (agri-tech, hospitality, construction, healthcare) will become a natural payroll distributor. Players such as <a href="https://www.checkhq.com/">Check</a> and <a href="https://www.salsa.dev/">Salsa</a> explicitly target these actors as channels.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FURixQxVErbJ4h_CRhRIoA.png" /><figcaption>Overview of Check’s interface</figcaption></figure><h4>#6— <strong>AI-native Payroll for startups &amp; SMEs</strong></h4><p><em>How to fully automate the HR/Finance back-office (payroll, taxes, benefits, compliance) for teams of 10 to 500 people?</em></p><p>Unlike the first wave of payroll SaaS (which digitized the spreadsheet), these players build their products around autonomous AI agents capable of orchestrating complete workflows. YC is simultaneously supporting several players in this segment (<a href="https://www.warp.co/">Warp</a><strong>, </strong><a href="https://www.every.io/">Every.io</a><strong>, </strong><a href="https://centralhq.com/">Central</a><strong>)</strong>, a sign that “autonomous payroll” is considered one of the most promising verticals of agentic AI.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xxdeBKU1vtH-kKhWYQvktw.png" /><figcaption>Overview of Check’s interface</figcaption></figure><h4>#7 — Back-Office All-in-One</h4><p><em>Why not using payroll as the gateway to a fully integrated back-office?</em></p><p>The logic is simple: the employer must run payroll anyway; it might as well become the hub to which banking, accounting, tax, benefits and time management connect.</p><p><a href="https://www.every.io/">Every.io</a> summarizes the thesis: “Why should a founder juggle 6 different tools for tasks that could run in the background?” In 2026, the company went further with native AI agents (AI CFO, AI Bookkeeper, AI CHRO), aiming to fully automate the startup back-office.</p><p>But this trend is not limited to tech startups:</p><ul><li>US player <a href="https://fingercheck.com/">Fingercheck</a> applies this logic to the “deskless” employee market (construction, hospitality, healthcare), which represents 80% of the global workforce. Created in 2013, the company raised a major round of $115M in late 2024.</li><li><a href="https://gusto.com/">Gusto</a>, for its part, is targeting SME, with over 400 000 clients! Valued over $10B, the company is now expanding its product through acquisitions, notably into retirement plans (see <a href="https://www.cnbc.com/2025/08/27/gusto-retirement-guideline-deal.html">Guideline’s acquisition</a>).</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*mJIUH0ksbMYrDknfSOTX9Q.png" /><figcaption>Overview of Gusto’s interface</figcaption></figure><h4>Food for thought</h4><p>What we see is a function that is still partially manually operated, with a European regulatory deadline in 18 months, and a technology, AI agents, now mature enough to automate the entire chain. The question is no longer whether organizations must modernize their payroll. It is how fast, and who will be agile enough to not find themselves behind on the June 2027 deadline.</p><p><em>Special thanks to our readers, as well as Sixtine Granoux for the mapping. You are a founder building a company in the payroll or compensation space and you would like to reach out? Contact me at gf@educapitalvc.com or on </em><a href="https://www.linkedin.com/in/geraldinefillet/"><em>LinkedIn</em></a><em>.</em> 📥</p><p><strong>*Sources</strong></p><ul><li>EU Pay Transparency Directive — Ravio (complete guide)</li><li>EU Pay Transparency Directive — Safeguard Global (2026 update)</li><li>Pay Transparency Laws by State 2025 — Littler</li><li>Salesforce acquired Spiff for $419M — TechCrunch (May 2024)</li><li>Global Payroll Report 2024 — Remote</li><li>Sales Compensation Statistics 2025 — Everstage</li><li>AI trends in compensation 2025 — Beqom</li><li>Payroll Intelligence for the Agentic Era — KPMG 2025</li><li>Warp raises $18M Series A — DHRMap</li><li>Rain raises $75M Series B — Business Wire (April 2025)</li><li>Gusto $175M Série F at $10B valuation — CB Insights</li><li>Fingercheck $115M — Edison Partners (October 2024)</li><li>Ravio $12M Série A — EU Startups (May 2025)</li><li>Rippling vs. Deel — CNBC Series G $450M (May 2025)</li><li>CFPB EWA advisory opinion — Davis Wright Tremaine (December 2025)</li><li>EWA market growth — Fortune Business Insights</li><li>API-Based Payroll Tech Market Report 2026 — Research &amp; Markets</li><li>Paychex to acquire Paycor for $4.1B — Reuters (April 2025)</li><li>ADP acquires WorkForce Software for $1.2B — WSJ (October 2024)</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2f675a49e1f3" width="1" height="1" alt=""><hr><p><a href="https://medium.com/e-d-volution/payroll-in-the-age-of-ai-2f675a49e1f3">Payroll in the Age of AI</a> was originally published in <a href="https://medium.com/e-d-volution">Educapital</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Edtech & Future of Work — February 26]]></title>
            <link>https://medium.com/@educapital_vc/edtech-future-of-work-february-26-eaf41d7307f8?source=rss-5c4f305f8e64------2</link>
            <guid isPermaLink="false">https://medium.com/p/eaf41d7307f8</guid>
            <dc:creator><![CDATA[Educapital]]></dc:creator>
            <pubDate>Wed, 18 Mar 2026 13:04:39 GMT</pubDate>
            <atom:updated>2026-03-18T14:07:35.846Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/430/1*nVbr4Zc6JtInL4BZNJtiyQ.png" /></figure><h3>Edtech &amp; Future of Work — February 2026 Snapshot</h3><ul><li>In February 2026, global funding in EdTech and the Future of Work raised <strong>c.$447m,</strong> compared with $441m raised in February 2025, <strong>meaning a 1% increase</strong>. Year-to-date, the sector raised a <strong>total of $817m.</strong></li><li><strong>North America and Europe were the most active regions this month</strong>, together accounting for the majority of deals and capital raised.</li><li><strong>Night</strong> <strong>and</strong> <strong>Naboo</strong> closed the largest rounds of February at<strong> $70M Series B each: </strong>Night to expand its gamified language-learning platform for children, and Naboo to help large organizations automate their event spending globally. Other notable deals include <strong>Humand ($66M Series A, USA)</strong> for international expansion, <strong>Subject ($28M Series B, US)</strong> to scale AI-powered K–12 Curriculum and Online Learning, and <strong>Kombo ($25M Series A, Germany)</strong> to build the infrastructure layer for people data.</li></ul><h3>Some notable funding rounds around the world:</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/860/1*NhnB-I72aPm24V55L24hVQ.png" /></figure><p><strong>M&amp;A Activity This Month</strong></p><ul><li>There were <strong>25 M&amp;A deals</strong> this month (VS 28 in January 2026).</li><li>This month, <strong>Valsoft</strong> announced the <strong>$7M acquisition of DigitalEd</strong>, a Canada-based cloud software company providing online education solutions for science, engineering, and mathematics courses. <strong>Sofatur</strong> also announced the <strong>$4M acquisition of SchoolMouv</strong>, an online education platform offering video-based courses.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/400/1*xXijFau9aESuUu_TuTtNRQ.png" /></figure><p><strong>What’s new in AI — February 2026</strong></p><ul><li><a href="https://fr.finance.yahoo.com/actualites/deepseek-s-appr%C3%AAte-%C3%A0-lancer-153258047.html"><strong>DeepSeek</strong></a> is set to launch its next-generation <strong>V4</strong> model in mid-February 2026, featuring a hybrid architecture and 1M+ token context window designed to outperform competitors in complex coding and reasoning tasks.</li><li><a href="https://techcrunch.com/2026/02/19/googles-new-gemini-pro-model-has-record-benchmark-scores-again/"><strong>Google</strong></a> released <strong>Gemini 3.1 Pro</strong> in preview, achieving record benchmark scores including a <strong>77.1%</strong> on ARC-AGI-2 (more than double the reasoning performance of Gemini 3 Pro).</li><li><a href="https://claude.com/blog/cowork-plugins-across-enterprise"><strong>Anthropic</strong></a> launched <strong>Claude Cowork</strong> plugins, enabling enterprises to build private marketplaces for AI agents that can orchestrate tasks across Excel, PowerPoint, Salesforce, and Google Workspace.</li><li><a href="https://www.perplexity.ai/fr/hub/blog/introducing-perplexity-computer"><strong>Perplexity</strong></a> introduced <strong>Perplexity Computer</strong>, a unified management platform that orchestrates <strong>19 different AI models</strong> to handle end-to-end projects from research to deployment.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*TpLO5LPlyAvewhTNY-TXPA.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=eaf41d7307f8" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Edtech & Future of Work — January 26]]></title>
            <link>https://medium.com/@educapital_vc/edtech-future-of-work-january-26-39e33457b64d?source=rss-5c4f305f8e64------2</link>
            <guid isPermaLink="false">https://medium.com/p/39e33457b64d</guid>
            <category><![CDATA[tech]]></category>
            <category><![CDATA[vc]]></category>
            <category><![CDATA[productivity]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[fundraising]]></category>
            <dc:creator><![CDATA[Educapital]]></dc:creator>
            <pubDate>Tue, 17 Feb 2026 14:10:57 GMT</pubDate>
            <atom:updated>2026-02-17T14:10:57.446Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/526/1*ax9BHNm2joP9dRAp40Vbgw.png" /></figure><h3>Edtech &amp; Future of Work — January 2025 Snapshot</h3><ul><li>In January 2026, EdTech and Future of Work companies raised <strong>c.$396m</strong>, above the <strong>c.$313m recorded in January 2025</strong>.</li><li>Funding activity this month was heavily concentrated in <strong>Europe</strong>, which captured 54<strong>%</strong> of all capital raised. This was largely driven by a standout late-stage round for <strong>Preply ($150m Series D). North America, comes second </strong>with<strong> </strong>Series B rounds for <strong>Benepass</strong> ($40m) and <strong>Compa</strong> ($35m) both Future of Work companies.</li></ul><h3>Some notable funding rounds around the world:</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/710/1*THD173mfP_HMX1w8oLT0NQ.png" /></figure><h3>M&amp;A Activity This Month</h3><ul><li>28 exit deals were recorded in EdTech and the Future of Work in January 2025.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/542/1*HykmQ_AOh8fcGAioFH6ENg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*nHemV7HX8csmtspsiRbWaQ.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=39e33457b64d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Preply, an insider story]]></title>
            <link>https://medium.com/@educapital_vc/preply-an-insider-story-c2c6b6252b60?source=rss-5c4f305f8e64------2</link>
            <guid isPermaLink="false">https://medium.com/p/c2c6b6252b60</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[vc]]></category>
            <category><![CDATA[tech]]></category>
            <category><![CDATA[unicorns]]></category>
            <dc:creator><![CDATA[Educapital]]></dc:creator>
            <pubDate>Tue, 17 Feb 2026 14:09:08 GMT</pubDate>
            <atom:updated>2026-02-17T14:09:08.788Z</atom:updated>
            <content:encoded><![CDATA[<p><em>In 2020, as COVID was about to hit, we invested in a small </em><strong><em>Ukrainian startup’s Series A: Preply</em></strong><em>. Last month, it announced its </em><strong><em>$150m Series D</em></strong><em>, led by </em><strong><em>WestCap</em></strong><em>, experts at scaling marketplaces like Airbnb, valuing the company at </em><strong><em>$1.2bn</em></strong><em>. For us at Educapital, this milestone is tangible proof of our core thesis: </em><strong><em>technology can improve education outcomes at scale</em></strong><em>. Here is how we built a conviction that we needed to invest in Preply and the story since.</em></p><h3>Why we invested in 2020</h3><p>With hindsight, one could assume investing in Preply was an easy decision at the time. <strong>It wasn’t. </strong>The world was in turmoil due to Covid and everyone was starting to operate remotely.</p><p>And yet, we ultimately decided to go for it for three reasons :</p><h4>1) A resilient team with boundless ambition</h4><p>By the time we met the team, they had already proven their resilience: starting to scale internationally from their Ukrainian office with <strong>less than $50,000</strong> in funding for the first couple of years before raising venture funding, <strong>not paying themselves for long periods</strong>, building an international marketplace from a region that has lived with <strong>conflict since 2014.</strong></p><p>Just as important: the founders’ ambition was huge. During our investment committee, when asked what their goal for Preply was, Kirill simply answered: “<strong>build a $10 billion company”.</strong></p><p>Kirill is one of those rare founders who can <strong>both articulate the full vision and knows the business intimately </strong>down to every detailed metric.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*MVXxmoy0V1gREFnOw5zjrw.png" /><figcaption>Serge Lukyanov, Dmytro Voloshyn, and Kirill Bigai</figcaption></figure><h4>2) A massive — but inefficient — market</h4><p>For hundreds of millions of people worldwide, learning a new language isn’t just a hobby or an academic requirement, it’s <strong>a way to improve their opportunities in life.</strong></p><p><strong>Crucially, it’s also a huge $85bn/year market.</strong></p><p>Duolingo, despite its popularity, isn’t enough to become fluent in a language. Until recently, the dominant path to real progress was still <strong>in-person teaching, for groups or individuals</strong>. It was <strong>expensive </strong>and finding good tutors for the language you wanted to learn and <strong>scheduling classes could be difficult.</strong></p><p>By moving learning from classrooms and homes to a remote model, Preply <strong>unlocked a global supply of teachers and learners</strong>. And its users were highly motivated,to this day, most of Preply’s learners use it for key milestones in their life:</p><ul><li><strong>Accelerating their career,</strong></li><li><strong>Emigrating and integrating into a new country,</strong></li><li><strong>Preparing for and passing exams</strong> for work or study,</li><li><strong>Building relationships</strong> with friends, partners, or family.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/786/1*tBzU_uCU5M1P_c7FyTMXmg.png" /></figure><h4>3) A scalable model for language learning</h4><p>The traditional language learning market didn’t scale well because it was physical by default: <strong>local supply, local demand, limited matching</strong>, scheduling and language options.</p><p><strong>Preply flips that dynamic</strong>. It enables teachers around the world to access a vast pool of learners and, in many cases, to make teaching their full professional activity. As the marketplace grows, it benefits from economies of scale: better matching, more specialization, more availability, and limited marginal costs at the platform level. Under the hood, Preply is built on a sophisticated tech infrastructure designed to maximise their marketplace liquidity: their matching algorithm analyzes hundreds of thousands of signals from budget, to time zones, availability and learning style to pair students with tutors in seconds.</p><p>This was already demonstrated by <strong>impressive traction at the time, with very optimized marketing costs. Preply are true masters at building an incredibly powerful programmatic SEO engine: </strong>they capture global demand organically by dynamically generating landing pages for millions of specific search combinations.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/726/1*dYy-2RtdnATxMCA_ou2ifw.png" /></figure><h3>A perfect fit for our thesis : improving education outcomes through technology</h3><p>As an impact fund, <strong>our final criteria to invest was assessing the fit with Educapital’s core mission : leveraging new technologies and business models to improve education outcomes.</strong></p><p>We built Educapital on the conviction that delivering high quality and inclusive education is not only a goal for society, but also sound business. We have built a <strong>proprietary impact framework</strong> to assess this:</p><p><strong>Reach: </strong>Preply had the potential to serve anyone in the world. It currently has an astonishing 800.000 learners on its platform<strong>.</strong></p><p><strong>Inclusion: </strong>Preply expanded access to language learning by using remote delivery and a global tutor base, bringing high-quality 1:1 lessons to many learners at more affordable price points.</p><p><strong>Outcomes: </strong>users had access to high quality individual sessions, augmented by technology, the gold standard in education. Preply now publishes <a href="https://preply.com/en/blog/preply-leanlab-2025-efficiency-study/">efficiency studies with third-parties</a> to validate their results. But beyond that, it provides stable employment for tens of thousands of tutors, and life opportunities for its learners.</p><p>Crucially for us, there is<strong> strong alignment between business success and impact.</strong></p><h3>Preply’s journey</h3><p>Since we invested, the team has been <strong>laser-focused on realizing the same vision.</strong> They are now the undisputed leaders among language learning marketplaces, and one of the largest Edtechs in the world. They make no secret of their ambition to dominate the market and eventually IPO.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*IGl4SPjXNQPG7xHGAjfetg.jpeg" /></figure><p>There were a lot of key moments along the way:</p><ul><li>Their Series C happened the day before Russia’s invasion of Ukraine. Despite the war, they kept their office and engineering teams in Kyiv<strong>.</strong> They managed to keep the service running at 100% uptime despite the disruptions,</li><li>They moved their global headquarters to New York and appointed C-level leaders from Airbnb and Google, to keep scaling in the US,</li><li>They transformed the business from one-off spending to a <strong>subscription-based service, drastically improving retention</strong>,</li><li>They <strong>accelerated on B2B</strong> which is now a significant part of their business,</li><li>They kept a lean mindset, stemming from the company’s tough beginnings, enabling them to <strong>reach profitability in 2025.</strong></li></ul><p><strong>We have reinvested multiple times since. Those decisions were much easier…</strong></p><h3><strong>AI at the core to serve learning</strong></h3><p>Preply has been very active in using AI, integrating in its <strong>core matching algorithm</strong> early on. But it has<strong> no plans to replace teachers</strong>. Rather, it aims to deliver the best learning possible by building an <strong>AI copilot </strong>for them : giving teachers the best tools and automating routine tasks like lesson summaries and homework generation.</p><p>There is a lot more coming on the roadmap, but we’ll let them announce it in time.</p><h4><strong>Conclusion</strong></h4><p>Preply becoming a unicorn is a headline. But the story underneath is what matters to us. It is a story of <strong>Ukrainian resilience</strong>, of a<strong> B2C model that conquered the world</strong>, and of a <strong>team that never lost sight of their mission</strong>.</p><p>We are incredibly proud to support them as they continue to democratize access to learning, proving once again that the best way to predict the future of education is to invent it.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c2c6b6252b60" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>