Atomico

It's State of European Tech week which also means it's Slush! 🎉 Tuesday 19th, midnight GMT - #SOET24 report and documentary launch 🎤 Wednesday 20th, 3pm EET - Catch Tom Wehmeier and Sarah Guemouri on the founder stage, with the report highlights 🎦 Thursday 21st, 9am EET - Watch the documentary on the big screen at the founder stage And meet our investment team on the ground with Laura Connell, Luca Eisenstecken, Hillary Ball, Andreas Helbig, Pranav Vangala, Lisa Sorg, Amanda Hultman and Miguel Rodrigo Pérez representing Atomico. Sign up to our newsletter for updates https://lnkd.in/eGRENwCU #europeantech #venturecapital

AI agents are getting easier to build, deploy, and swap in and out. That’s great for innovation but tough for defensibility. If the core tech is widely accessible, where does long-term value sit? I think we’re about to see a rebundling of systems of record like HRIS, ERP, and CRM, as owning the source of truth becomes more valuable than any single AI solution on top. Companies that consolidate and structure data effectively will be the foundation for the AI agent ecosystem. But the real winners won’t just own the system of record, they’ll bundle everything: the system of record, the automation platform, their own AI agents, and a marketplace for third-party agents. Controlling the full stack, from data to execution, will be the key to long-term defensibility. Some companies are taking this to the extreme. As Sebastian Siemiatkowski shared last week, instead of relying on traditional solutions Klarna is building tools on top of their own internal system of record data store, replacing external platforms entirely. Few companies can pull this off, but it raises a question: Will more try as developing tools in house gets ever easier? And as AI agents take on more execution, the system of record concept will also evolve, capturing agent-driven workflows, automation history, and execution patterns. This could create a new kind of lock-in, where switching platforms isn’t just about migrating data but also retraining AI agents that have learned an organisation’s workflows. Are we heading toward a future where the next great platforms control all layers? Or will modular, API-first approaches win out instead?

Tune into How I Built This with Guy Raz this week for a bonus episode with Atomico founder Niklas Zennström discussing his founder journey from Kazaa to Skype. The highs, the lows, the lawsuits - it's a nail biting ride. Here are a few highlights: 💸 Trying and failing to raise money for Kazaa from Amsterdam in the wake of the dotcom crash, and eventually bootstrapping to massive viral success 🤝 How a trip to the US to collaborate with the recording rights organisations quickly turned into a lawsuit 🏃 Leaving the States to avoid being served papers, and accepting there was no immediate way to generate revenue from Kazaa, despite huge user take up 📞 Using this underlying tech to break the telecoms monopoly and build Skype 💵 Facing yet more funding issues in the journey to create Skype. Their first, company saving, investment was a small cheque from legendary Silicon Valley VC Bill Draper 🎤 How even with fears of significant user friction - both users had to download the service, they had to connect a microphone, and be at their desk to make a call - viral take up of the new Skype was once again HUGE 📈 Skype hitting 100m subscribers inside 2.5 years and eventually selling to Microsoft for $8.5bn - Europe's largest tech M&A at the time Thank you to Guy Raz and the team NPR for having us on the show! https://lnkd.in/ebeKjYey

The AI agent pricing paradox: vendors are selling next-gen technology through last-gen pricing models. When RPA emerged, we locked into annual contracts priced based on the cost of the employees they replaced. Now AI agents offer unprecedented scalability—yet we're still pricing as if they're conventional employees with fixed costs. As Aaron Levie recently highlighted on Y Combinator's Lightcone podcast, the transformative potential lies in shifting "work" from fixed to variable costs. Why substitute the fixed cost of an SDR with one AI agent of a similar annual price, when you could pay for and deploy 100 agents that generate 1,000 qualified leads in a week, and then switch that cost off until you need it again? The misalignment is striking: AI's core value proposition is eliminating fixed overhead in favour of dynamic, on-demand capability—yet pricing models preserve the very constraints the technology aims to eliminate. Two things in the ecosystem will need to crack: 🔹 Founders: How can your pricing best reflect the actual value delivery mechanism of AI agents? 🔹 Investors: How should you think about revenue quality metrics when maximum customer value comes from usage-based models with variable demand? Does investor preference or CFO demand for predictable contracts ultimately prevail, or will pricing evolve toward the flexibility that makes this technology truly revolutionary?

Really great to host Daily's first Voice AI Meet Up in London at Atomico together with ElevenLabs and Google DeepMind. We saw some very cool and powerful demos from Pipecat, ElevenLabs' new ASR model, Gemini Flash's multimodal capabilities and spoke about conversational AI, agent-to-agent interactions, ethics & safety and more. Big thank you to all the speakers and attendees! Kwindla Hultman Kramer, Jonathan Taylor, Angelo Giacco, Juliette Love, Ian Ballantyne

Atomico's Alvaro Campos, Victor Jolly and Caroline Chayot were in Barcelona this week for 4YFN and to co-host a fantastic dinner with the team at Yellow. It was an evening full of energy and talent, reflecting our ongoing commitment to supporting early-stage founders and innovative operators across Spain and Southern Europe. We’re especially grateful to our portfolio founders Job&Talent's Felipe Navío, Factorial's Jordi Romero and Abacum's Jorge Lluch for joining us and sharing their insights. A big thanks to Adam S. Lasri for helping us convene such a stellar group.

'People are vastly underestimating what's possible with this next generation of AI in terms of productivity and efficiency gains' Thank you to James Fitzsimmons and Clif Marriott for inviting Atomico Partner Laura Connell to discuss the shifting paradigm for AI with Synthesia's Steffen Tjerrild, causaLens' Darko Matovski, PhD and Eric Sheridan at the Goldman Sachs Disruptive Tech Symposium

🇪🇺🇸🇪 Last week at TECHARENA, I had the chance to discuss the future of European tech with some incredible panelists — Mimi Billing, Laura Modiano, and Maxine Rior. One key takeaway? The biggest crisis we face in Europe is a crisis of confidence. Here’s why I believe Europe is stronger than ever: 💡 AI is far from just hype — beyond the headlines, real substance is driving investment and innovation. Industry-specific, vertical AI is transforming sectors like food, construction, manufacturing, and heavy machinery training — where Europe has deep expertise. 🌍 Europe has unmatched technical talent & sovereign strength — we don’t need to play catch-up; we are already leading in key areas. From connectivity to FinTech to manufacturing, Europe is uniquely positioned to shape the next wave of game-changing startups. 🚀 We are building category-defining companies (Klarna, Revolut, Celonis, HealX, Hinge Health) — too often, the cliché is that European founders "think small." The reality? Many of our startups are already scaling globally with both European and US backing. The ambition is there — we just need to double down on our belief. 🫡 There’s never been a better time to build in Europe. Let’s shift the narrative from self-doubt to confidence — the next wave of global tech leaders will come from this ecosystem.

Introducing Omega, the first manufacturable chipset for photonic quantum computing. Designed by PsiQuantum and manufactured at GlobalFoundries, this chipset contains all the advanced photonic components required for million-qubit-scale quantum computers. Featured in a newly published paper in Nature Magazine, every photonic component necessary is demonstrated with beyond-state-of-the-art performance. The paper shows high-fidelity qubit operations, and a simple, long-range chip-to-chip qubit interconnect – a key enabler to scale that has remained elusive for other technologies. The chips are made in a high-volume semiconductor fab, representing a new level of technical maturity and scale in a field that is often thought of as being confined to research labs. Learn more here: https://lnkd.in/evBsEn6U

🇪🇺🚀 NEW REPORT: Accelerating Europe: The State of European Innovation and Why It Matters. The opportunity in Europe right now is unprecedented. Europe’s scaleup potential has never been greater. But the right conditions are crucial. What Europe does next matters. This report, in partnership with Dealflow.eu, examines four crucial dimensions: • Unpacking Europe’s GDP gap with the US • Big tech's impact • Europe's VC ecosystem - the strengths and gaps • Pathways to unlocking European growth, with insights from the Draghi report This research provides context, fresh perspectives, and empirical evidence to enrich the ongoing conversation about Europe's tech future. Particular thanks to contributors and collaborators Tom Wehmeier (Atomico), Roxanne VARZA (STATION F), Andreas Klinger (Prototype Capital 🦾), Paul Murphy (Lightspeed), Philipp Herkelmann (The EU Inc Petition), Nathan Benaich (Air Street Capital), Michael Jackson (Multiple Capital), Lucien Burm (Dutch Startup Association/Prompting Europe), Yoram Wijngaarde 🙏 📊 Download the full report below. And stay tuned - Part Two drops on Monday  #innovation #productivity #economicgrowth #europe #startups #investment