Silicon Roundabout Ventures

It's out! Our recent investment into precision-strike drone startup built by Airdog's exited founders 📰📣 So much empty talk about defence tech around. Mostly to back companies forced to shy away from the real problem. However, very few are focusing on true warfare issues like it's happening in Ukraine or the the middle east at the moment. Not this one. Enter is a true game-changer alert: Origin Robotics's Precision Drones! A Latvian startup crafting reusable, autonomous drones. War is evolving. Smaller nations need affordable defence solutions. These drones deliver munitions cost-effectively. Backed by deals with two NATO countries, Origin is already making waves (big shout to Mike Butcher ✍️ and TechCrunch for the work behind the round announcement!). Founded by Agris Kipurs and Ilja Nevdahs, the brains behind Airdog, Origin leverages over a decade of drone expertise. Their flagship, BEAK, is an ISR drone with precision guidance. It's tough, portable, and built for modern warfare. Instead of self-destructing, BEAK delivers munitions while remaining intact. This reusability slashes costs significantly. Imagine a flying Javelin—not just a drone. Precision strikes at a fraction of the cost. Origin's tech empowers smaller democracies to defend themselves efficiently. With €2.4 million raised and EU support, Origin's total funding hits €4 million. We at Silicon Roundabout Ventures are proudly investing in this transformative venture alongside the round leads over at Change Ventures (Andris K. Berzins) Join the conversation! How do you see precision drones reshaping defence strategies? Share your thoughts below!

Announcing the 4th and final judge for The Deeptech Demo Day 2024 ! 🦄⚛️ Julia Hawkins, General Partner at one of the most successful seed funds globally: London-based LocalGlobe! (now spanning multiple funds across seed and growth & a physical space under the brand Phoenix Court) Julia focuses on health and deeptech. 🕵 Her investments include: Oxford Nanopore Technologies (LSE:ONT), Barinthus Biotherapeutics (formerly Vaccitech, Nasdaq: BRNS), AccuRx, Inc, YuLife, CoMind, Vira, Gilion and Spore.Bio. 🚀 Her mission is to back ambitious, thoughtful and diverse founders to transform healthcare from scarcity of clinicians and appointments to ease of access through technology and to enable new and better treatments to reach patients. 🧪 About her investment drive, Julia says: "My daughter was diagnosed with a rare condition which meant that I had to navigate several health systems and I found it really difficult to get to a diagnosis. But I also saw that there was a lot of promise in terms of both how healthcare could be delivered, if better digital tools were used, and how technology could help to get to better therapies." 🌍 Julia’s centre of gravity is London with The Francis Crick Institute on Phoenix Court’s doorstep and Google DeepMind, UCLH Charity and Royal Free London NHS Foundation Trust nearby. 🇸🇪 🇬🇧 🇪🇺 🇮🇱 🇺🇸 ❤️ She’s Swedish and is excited about fantastic Nordic founders with global ambitions, and is seeing phenomenal healthcare talent in Israel and Boston. 💡 What she looks for from founders: "I really like founders that can help me understand their unique perspective – whether it be life experiences, or previous companies that they've built or been associated with – that have led them to the insights that they have. (...)" 👩💻 Julia’s career began at Goldman Sachs, but passionate about startups, she joined Last.fm in 2007 as Business Operations Manager – the ground-breaking music startup’s first generalist hire. In what became one of Europe’s early landmark deals, Last.fm was acquired by CBS. Julia then joined BBC Worldwide, the corporation’s commercial arm, as Digital Strategy Manager, and worked with all product teams from news to games, before joining Universal Music UK, where she helped launch its corporate venture arm. Julia and I co-invested on a really exciting founder together not too long ago, so I'm truly grateful she decided to join The Deeptech Demo Day and share her knowledge and experience with the other judges. 📅 Now that the 4 judges have all been announced: let the best team win! See you on November 7th, register to attend live via zoom. ➡️➡️➡️ Registration website linked on The Deeptech Demo Day page. ------------------------ Powered by Silicon Roundabout Ventures & Frontier Deep Tech (Dr Cristina Esteban Pardo) 😍 Thank you to our wonderful sponsors Mishcon de Reya LLP and OVHcloud -------------------------- #deeptech #venturecapital #angelinvesting #europeantech #demoday

Finally research showing what we all knew in computer science: Mainstream AI (LLMs) has HUGE limitations. Imho it won't bring us AGI. 🚀 Breaking Down AI's Math Skills: A Reality Check Large Language Models (LLMs) are the rockstars of AI, but can they really reason? 🤔 Apple engineers put them to the test with GSM8K, a math benchmark designed for grade-school questions. Results? Not as impressive as you'd think. 🔍 Key Findings: 1. Performance Drop: Change a number in a question, and watch LLMs struggle. Their accuracy drops significantly—up to 65% when irrelevant info is added. The research are pretty clear: "There is just no way you can build reliable agents on this foundation, where changing a word or two in irrelevant ways or adding a few bit of irrelevant info can give you a different answer". 2. Complexity Struggles: Add more clauses to a question, and performance tanks. LLMs aren't doing real logical reasoning, instead they're mimicking patterns from their training data: "We found no evidence of formal reasoning in language models," the new study concluded. The behavior of LLMS "is better explained by sophisticated pattern matching" which the study found to be "so fragile, in fact, that [simply] changing names can alter results." 📉 3. Why It Matters: These findings highlight a major gap in AI's ability to handle real-world problem-solving, especially in math. And it shows that there are plenty of non-LLM based AI infra opportunities for non-mainstream VCs like Silicon Roundabout Ventures and our 120 friends signed for The Deeptech Demo Day: Personally I'm very excited by new hardware solutions that combine software in a way that creates alternatives to the combo LLM+GPUs. And of course there is even more that can said about non-transformer based models yet to be popularised by a new wave of AI startups. 🔧 The Road Ahead: We need better evaluation methods and more research to build AI that can truly reason like humans. Until then, don't rely on LLMs for your math homework, as the paper says: "Recent advancements in Large Language Models (LLMs) have sparked interest in their formal reasoning capabilities, particularly in mathematics. " "While the performance of LLMs on GSM8K has significantly improved in recent years, it remains unclear whether their mathematical reasoning capabilities have genuinely advanced, raising questions about the reliability of the reported metrics. " Check out the paper on arXiv: 07 Oct 2024 ─ GSM-Symbolic: Understanding the Limitations of Mathematical Reasoning in Large Language Models #deeptech #ai #ventuerecapital ___________________________________ Like this content? I’m Francesco Perticarari, Deeptech Venture Capitalist focused on European pre-seed. I share my experience and insights building my soloVC firm. Follow form more 🔔

⏰ Deadline today: Calling all European deeptech founders! 📣 This is for you 👇⚛️🦾 Today's the last chance to apply for The Deeptech Demo Day 🚨 This is not your average pitch fest. Some of the most ground-breaking startups from all over Europe have applied: → Supported by over 120 VCs. → 3 top-notch fund General Partners confirmed as judges. → 10 trailblazing finalists will showcase live. Powered by Silicon Roundabout Ventures and Frontier Deep Tech ( Dr Cristina Esteban Pardo) 🚀🚀🚀 Every applicant gets exposure to all VCs involved. Last year, we made over 100+ founder-VC introductions. With the VCs pitching for meetings to the most relevant companies. This year, we're nearly doubling our VC support. From local Europeans to US and global investors. From first cheque soloGPs to huge funds backing series A and beyond. This is the defining moment for European deeptech. If you're a deeptech entrepreneur, this is your stage. Time is ticking. Today is the last day to apply. Don't let this opportunity slip. Whether you're actively raising or not, this is your chance to showcase your technology and brand to the world. Ready to make waves in the deeptech world? Open and free for all pre-seed to pre Series A deeptech founders in the whole of Europe. Like and tag someone who should to help founders and the ecosystem 🙏👇 Application link on the comments below and on The Deeptech Demo Day page --- Sponsored by: Mishcon de Reya LLP and OVHcloud 🙏

Is VC activity related to how many people invest in stocks and shares? Then Europe doesn't need more money: it needs a new CULTURE 💸🧑💻 Over the weekend I went into an investment research binge and found a striking relationship between VC investment and the population rate of investment in the stock market. Here what I found: It all started by reading in the FT that as a response to the raising cost of living, Europeans started to hoard cash. US folks seem a lot more relaxed about their future, with spending (and investing) continuing on a buoyant pace. I started to look at history and trends at what people do with their money, in general. And it turns out that data shows what's quite evident to a European random guy from a little countryside village like me: We just don't have the investment culture and risk appetite the average US person has. In my immediate family, for instance, I'm the ONLY one who's got funds invested in stocks and shares. Whilst I couldn't find any specific research linking the two aspects, there is a striking similarity between % of the population invested in stocks and shares and the % of global VC investment in a given region. Last year, globally, about $320 Billions were invested in Venture Capital. With the UK and Europe taking up about 20% of that. But according to Sifted, a worrying 30%+ of that money comes from government LPs. So when we say we want Europe to compete with the US on VC funding, growth funding, IPOs... ...Why everyone seems to worry just about the headline figure of dollar amount? Am I the only one to see this as unsustainable? Look at China: the government pumped money into startups and US VCs flocked to the raising economy to make it the 2nd largest VC market globally some years ago. Now the government is more concerned about control than spurring startup activity and globalisation is falling apart. Result? China's startup investment is in free fall. Europe has one of the largest share of global savings but nor private individuals nor pension funds seem to care about equity. It's crazy that as private and public equity in the last 24 years created trillions of wealth, only 20% of the European population took part. Contrast that to the US where over half of the population is invested in the stock market and its global share of venture capital investment is also over 50% Unrelated? I don't think so. So maybe the question we should ask is not how to create more growth funds in Europe. But: how do we radically change the underlying culture? #deeptech #familyoffices #limitedpartners #venturecapital #angelinvesting I'm a soloVC investing in deeptech & building my firm and community in public. 📢 ❤️ & Follow+ to keep up with the journey 🛣️

Your lawyer is killing your company (or portfolio) ─and they don't even know. If you are in venture capital, as a company builder or investor, chances are you want to spend time in the action of growing your business, not dealing with corporate law. And many of you will probably assume lawyers are an unavoidable expense we take on when contracts need to be drafted or enforced. Then you reach out to someone with good reviews and that has experience drafting contracts and that's reasonably priced. Wrong. I made that mistake myself. And in the last year alone I witnessed lawyers screw up companies and investors alike: ↪️ Once they downright killed a company by pouring gasoline on a spark started by founder conflict and creating a bonfire. It's ok (for them), they got their fees upfront. ↪️ In another instance an HR law firm created contracts that forced a startup to pay out a corporate-level severance to a non performing employee who was in the job less than a year. ↪️ In another case they took fees worth 5% of the whole round (tens of thousands) for a mere seed round that could have been done with SAFEs. ↪️ And finally, the most ubiquitous: the lawyer that implements what you want during a funding round and drafts away the contracts you think you need. When in reality they should tell you your idea is not a good one for a VC backed company looking to to go far. Now here is the truth: There are only a few layers worth their salt in venture capital. People who don't care much about drafting contracts. You can do that on platforms these days and they may give you free templates if you ask. The good lawyers in the industry are advisors with 3 characteristics: 1) Venture capital transaction experience –and I mean real one, not helping a bunch of useless newbies do a pre seed. I mean taking several companies across the letters of the alphabet to exit. 2) Networks they are not afraid of using - if their LinkedIn is full of VC and Corporate connections across the US, Europe and other markets, that's a good sign, and the good ones will make intros for their clients 3) Long-term no bullshit approach: in most startup cases, the best approach is not to draft a contract at all, but maybe to refer to someone else or pick up the phone and speak to the other party, or to review a contract you drafted yourself. The right lawyer wants to represent you for your Series B+, M&A and IPO and charge you handsomely then. And not to be afraid to tell you your ideas are stupid and not market standard in VC land. Know how to recognise the above and pick the right partner lawyer (not just the firm) is how you can avoid others to (involuntarily) screw you out of ignorance. Your experience? #venturecapital #angelinvesting #startup

Tired of the "sorry we only invest in asset light software" BULLSHIT? Pitch grown up VCs now 🧑🔬🧑💻🔬⚛️ No, better: Have investors pitch YOU ☝️ >>> thedeeptechdemoday.com With 80+ Deeptech loving VCs partners ready to check out your engineering and science breakthroughs (and countless angels signing up for the finals)... ...it's time you apply for The Deeptech Demo Day 💸 Powered by Silicon Roundabout Ventures and Frontier Deep Tech and sponsored by Mishcon de Reya LLP and OVHcloud - this is Europe's only virtual demo day FULLY dedicated to the HARD-TO-BUILD, WORLD-CHANGING true tech startups. From lab spinouts to series A. If you are a European team looking to showcase your tech (whether or not you're raising at the moment), drop a quick application over and showcase immediately to our investors. PLUS if you get scored in the top 10, you'll get to present live (virtually) in November this year in front of all the VCs, corporate partners and angels attending. And this year we are doubling down with an invite only private event in London with exited founders, VCs and top corporate investors just for the finalists. What are you waiting for? Show the world European Deeptech founders are the absolute best. From computing to climate and from defence to medical tech, this is THE demo day just for you. Today announcing 20+ more VC partners now supporting us. A slide deck and 5' is all you need. Link in The Deeptech Demo Day page too. Follow it for updates! #deeptech #europeantech #startup

🔥[BIG NEWS]🔥 The 1st team we ever backed from this fund has just ANNOUNCED their SERIES A! 🍾 Anaphite got $13.7M in fresh equity funding to scale dry-coating EV battery technology baby 🎉👨🔬 What a tremendous effort from the founders Alexander Hewitt and Samuel Burrow and CEO Joe Stevenson 💪 Despite my little angel fund size, I couldn't be prouder to be doubling down ─this company can go on change the entire battery industry. And help us all un-fuck the planet as they do so 🌍 After being named a WIRED Trailblazer for 2024, it's time they now celebrate the real award: closing the first letter on the funding alphabet! What a milestone. Again, what an effort for these guys. Depending on which dataset you use, only 10-40% of seed funded startups ever get to raise an A round. And with over £10m in fresh funding in the bank and plenty of grants coming in, these folks can really go steamroll ahead and change battery manufacturing GLOBALLY. Forever. From tech and team perspective, it's great to see where they got to since our first meeting in Bristol before the seed round. Commercially, the company now is on the verge of announcing some major partnerships... But I guess you'll have to follow them to find out more on these 🤫📈 With the round closed, the company will now scale up their facilities to ship TONNES of their patented composite cathode material. As the A round lead investor mentioned: "Their technology is genuinely market-leading, and will help the world transition to an eco-mobility future. The team has already made breakthroughs in lithium-ion battery production, and their technology will be applicable to a wide range of new battery technologies, including sodium-ion and solid state." But wait. There's more: Anaphite’s proprietary technology will not only help spread the adoption of clean electrification, starting from EVs. It will also help Western battery companies and automakers onshore their battery development in the decades ahead. Why I got so excited when I backed them and I'm now even more so? Because battery tech costs remain a major hurdle for large-scale EV adoption, typically accounting for up to 30% of the total vehicle cost. And with Anaphite unique use of graphene chemistry and their proprietary replacement of wet-coating processes used in battery manufacturing they can save about 30% of the energy usage and 15% of factory floor space. This approach can MASSIVELY reduces the cost AND environmental impact of making a lithium-ion battery. Better batteries, with better range/specs AND cheaper = wider, cleaner adoption. Yes. Like a win-win-win kind of deal. What's not to love? ❤️ It's all on you guys: congrats! #deeptech #venturecapital

OpenAI sold you out. Literally. I'm cancelling my subscription. And the 2 are not even related. This week I hosted an event at the Italian Tech Week, where Sam Altman joined a fireside chat with Exor's John Elkan. Serendipitously, I also read a grat chronological account by Vox on how Sam Altman has single-handedly turned a (sort of) social enterprise into a possible global security and climate risk. And a possible money scam too. What I find worrying is not that a company or its leader seeks to make profit. I'm a venture CAPITALIST myself. And the founder of my own firm. Making money is how I make my living. But OpenAI took advantage of a (semi)non-profit status to advance technology (open sourced by others btw -i.e. Google) only to shift its agenda without any accountability for the consequences. Theses days OpenAI is reportedly pitching for 1 Trillion investment from oil money (Saudis etc) to bankroll a technology that's already accounting for a significant chunk of the global energy consumption. Folks on LinkedIn laugh about the EU wanting to regulate AI. But now even Americans and their publications are worrying. Even Musk, who's normally for extreme versions of capitalism. I now think it's a justified worry. When OpenAI was cofounded in 2015 the budding research lab introduced itself to the world with these words: "OpenAI is a nonprofit artificial intelligence research company. Our goal is to advance digital intelligence in the way that is most likely to benefit humanity as a whole, unconstrained by a need to generate financial return." OpenAI was a nonprofit controlled not by its CEO or by its shareholders, but by a board with a single mission: keep humanity safe. But! By 2023, the nonprofit board had grown so suspicious of the leadership that it tried to oust it's CEO. That didn't work. Earlier this year, OpenAI’s safety team imploded and quit the company. Now OpenAI will no longer be controlled by the nonprofit board. And Sam Altman is now getting equity worth billions. How could this be legal? If OpenAI does away with the profit cap, it would be redirecting a huge amount of money — prospective billions of dollars in the future — from the nonprofit to investors. Because the nonprofit is there to represent the public, this would effectively mean shifting billions away from the public. As magazine Vox notes: "it feels a lot like theft." And if somehow now OpenAI starts to onboard investors with dubious climate agenda and disregard for AI safety regulations, I wonder where all this will lead. Oh well, in the meantime maybe we should smile that their features and UX are losing pretty much against all their competitors: Claude is better for coding. Perplexity better for searching. Mistral for open sourcing and agent creation. So yes after all it's not consumer worries that are pushing me to unsubscribe from the premium OpenAI. It's just that they are falling behind despite all the money manoeuvres. Thoughts?

Back on the road to visit some startup labs yo ⚛️💸🥐☕ Love this part of a deeptech investor job. All but 1 of my fund investments had an element of lab visits. There is something uniquely insightful about seeing prototypes and meeting founders and early staff in their early day workshops. ...and humbling in being by far the dumbest in the room 🤓 #deeptech #startups #venturecapital #solovc #angelinvesting