European startups raised €5.1B across 290 deals in April 2026, with cleantech leading investment activity and the UK remaining the top fundraiser, despite a drop in overall capital.Click to read the rest of the news.

At Tech.eu, we keep track of the investment landscape with data-driven insights.   Our Tech.eu Insiders enjoy unlimited, exclusive access to all our content, including market-intelligence analysis, reports, articles, and useful insights on tech trends and developments.  But we know that a lot of folks interested in tech might not have the funds for a subscription. In response, we're offering compact versions of our monthly reports to all of our readers.  Our versions offer a glimpse into the valuable insights provided by our monthly reports, covering key investment trends, notable company activities, and emerging industry sectors. Download the March Tech.eu Pulse today.

Europe’s battery ambitions took a major hit with the collapse of Northvolt, once seen as the continent’s best chance to rival Chinese competitors and build a homegrown supply chain.  But since then, we’ve seen a more modular, ecosystem-driven approach with scaleups rebuilding momentum across the battery value chain, by tackling a specific bottleneck rather than trying to do everything at once. These are some of the key themes defining the space, along with companies to keep on your radar:  Battery recycling and circular economy The need to recycle batteries is quickly becoming one of the most urgent challenges in the energy transition. As electric vehicles, grid storage, and consumer electronics scale, so does the volume of lithium-ion batteries reaching end of life. Without robust recycling, this creates a growing waste stream and an escalating demand for raw materials that are already difficult to mine, expensive, carbon-intensive, and geopolitically sensitive. Cylib (Germany)   Cylib deeptech scale-up on a mission to revolutionise battery recycling and secure a circular future for Europe.  Born out of years of research at RWTH Aachen University, cylib has developed a holistic, water-based technology to recover all elements within lithium-ion batteries. The core challenge of Europe’s energy transition is its heavy dependence on imported critical materials. cylib’s technology addresses this by transforming end-of-life batteries – from electric vehicles, power tools, and energy storage systems – back into high-quality production materials. What makes their technology a game-changer is its holistic, water-based process, which achieves over 90 per cent recovery efficiency with 80 per cent lower emissions than traditional mining. Unlike many existing solutions, cylib can handle both NMC (Nickel Manganese Cobalt) and LFP (Lithium Iron Phosphate) battery chemistries at an industrial scale. The company has raised over €156 million. tozero (Germany) tozero focuses on recycling lithium-ion batteries by processing used or waste batteries to extract valuable raw materials, including lithium, graphite, nickel, cobalt, and manganese. In March, the company launched an industrial-scale battery recycling plant at the Chemical Park Gendorf in Bavaria, capable of processing more than 1,500 tonnes of battery waste annually and recovering critical materials such as lithium, graphite, and nickel-cobalt mixtures.  Check out our earlier interview with Sarah Fleischer, co-founder and CEO of tozero. What sets the company apart is its chemical (hydrometallurgical) recycling process, which allows it to recover a large share of these materials — often around 80 per cent or more — in a form pure enough to go straight back into industrial production.  The company has raised over €17 million R3 Robotics (Luxembourg) R3 Robotics  (formerly CircuLi-ion) is focused on automating the safe, scalable dismantling of end-of-life electric vehicles and their high-voltage components. When electric cars reach their end of life, taking them apart is still mostly manual, slow, expensive, and dangerous (especially due to high-voltage battery systems).  R3 Robotics' tech combines computer vision, AI, and specialised robotic tooling to automate the disassembly of lithium-ion battery packs, e-motors, power electronics, and other high-value electrified components. The system minimises human exposure to high-voltage hazards and delivers the cost structure and reliability required for industrial-scale operations.  The company has raised €28.5 million. For a batterytech deep dive, check out the EIC Scaling Club’s Market Roadmap: Batteries and Energy Storage. Battery R&D and manufacturing Europe’s battery maker scaleups are not only building batteries with more sustainable materials, but also focus on manufacturing locally, thus reducing Europe’s reliance on imported batteries and strengthening the region’s energy and EV supply chain. ElevenEs (Serbia) ElevenEs is the first company in Europe to develop a lithium-ion giga factory based on LFP blade-type cells.  LFP (Lithium iron phosphate) is a type of chemistry within lithium-ion batteries. While it’s a mature technology in China, Europe has been focused on more traditional options such as Nickel Manganese Cobalt (NMC) batteries.  By producing LFP (lithium iron phosphate) cells locally, ElevenEs offers a safer, lower-cost alternative that avoids scarce materials like cobalt and nickel while still meeting the needs of EVs and grid storage. Skeleton Technologies (Estonia) Skeleton Technologies develops high-power energy storage systems, primarily supercapacitors and hybrid SuperBattery technology, designed to deliver and absorb energy almost instantly.  Unlike traditional battery companies that focus on storing large amounts of energy, Skeleton is focused on power — how quickly energy can be released, captured, or stabilised. Its technology acts as a buffer within electrical systems, handling rapid spikes and drops in power that batteries alone cannot efficiently manage. The company’s systems are used across electric transport, energy grids, industrial machinery, and increasingly AI data centres.  In November 2025, Skeleton opened a €220M Leipzig SuperFactory to power Europe’s AI and grid stability Check out our interview with Skeleton Technologies  CEO Taavi Madiberk.  The company has raised €313.3 million. Battery energy storage  Battery energy storage (BESS) is a way to store electricity in batteries so it can be used later, rather than immediately when it’s generated. Batteries are uniquely valuable because they can both store and release electricity instantly, and they are becoming a key part of grid stability.  Sympower (The Netherlands) Sympower is an energy tech company that helps businesses use their electricity in a smarter way. It connects to assets like factories, batteries, or heating and cooling, and uses AI-embedded software to automatically decide when to charge or discharge them based on grid conditions.   By aggregating many batteries and flexible energy systems into a “virtual power plant,” Sympower can sell this flexibility into energy markets, helping stabilise the grid while generating revenue for its customers. When you look at Sympower through a battery lens, the company’s role becomes even clearer: it turns batteries into active, revenue-generating assets for the grid, rather than passive storage. Sympower connects to industrial batteries (and increasingly grid-scale storage systems) and uses its software to control when they charge or discharge based on grid needs. The company has raised $76.8 million. Check out our earlier interview with  Nikolas Samios, Managing Director, PT1, to understand how battery storage has become a rapidly evolving asset class. Battery intelligence Battery intelligence is emerging as a critical software layer in the energy stack, turning raw battery data into predictive insights that improve performance, safety, and asset value.  By analysing patterns in battery charging, discharging, and degradation, these systems enable operators to anticipate failures, optimise usage, and reduce downtime—an increasingly important capability as electric vehicles, grid storage, and industrial electrification scale. ACCURE battery intelligence (Germany)  ACCURE Battery Intelligence builds AI-powered software to monitor, analyse, and optimise battery systems. Its platform operates as a data and intelligence layer. It collects the huge volumes of data generated by battery management systems (BMS)— such as  voltage, temperature, and charge levels—and uses AI and physics-based models to turn that into actionable insights.  This allows operators to detect faults early, improve performance, and extend battery lifetime. This also helps prevent thermal runaway (battery fires), degradation, and inefficiencies by identifying problems before they become critical and recommending corrective actions.  With ACCURE, utilities, energy storage operators, and EV fleet managers can reduce risk, maximise uptime, and increase the economic value of battery assets. The company has raised $34.5 million. Companies mentioned in this article are members of the EIC Scaling Club, a curated community where 120+ European deep tech scale-ups with the potential to build world-class businesses and solve major global challenges come together with investors, corporate innovators and other industry stakeholders to spur growth. The EIC Scaling Club is an EIC-funded initiative run in partnership by Tech Tour, Bpifrance (EuroQuity), Hello Tomorrow, Tech.eu (Webrazzi), EurA and IESE Business School.

DeepL, the German AI translation startup, is cutting around 250 jobs, about a quarter of its headcount, saying it was moving to smaller teams so it is able to compete amid rapid AI advancements, its CEO said today. Posting on LinkedIn, Jarek Kutylowksi, CEO and founder, said the cuts at DeepL, which employs over 1,000 people globally, were a “deliberate structural choice about how DeepL needs to operate to remain a global AI leader”. He added: "The pace of AI has accelerated far beyond what was possible even just a year ago, changing what it means to operate effectively - including for a company like ours. "We are currently living through a massive structural shift in what work exists, who does it, and how many people it takes to do it well, and that shift is because of AI.” The Cologne-based AI translation startup, founded in 2017 and last valued at $2bn in 2024, is known for its AI text translation and writing tools. Kutylowksi said he and his management team had been reviewing how DeepL could best operate as a global AI firm amid fast improvements in AI. He added: "In practice, this means transforming how DeepL works from the inside out, with AI embedded into every layer of how we operate. "We are moving to smaller high-agency teams where AI handles the routine, so people can focus on what only humans can bring, like using our intuition, coming up with new creative ideas and seeing projects through from start to finish.  "AI systems will allow us to put more energy into the work that actually matters and move at a speed we haven’t seen before, leaving behind recurring roadblocks and day-to-day inefficiencies." The CEO did not say which specific areas of the business would be impacted by the cuts.

Cultivated meat pioneer Meatly, Europe’s first company to sell cultivated meat, has today announced that it has raised £10.4 million in Series A funding.  Since launching in 2022, Meatly has solved the key technical cost challenges facing the cultivated meat industry, accelerating the path to scalable, affordable production.  In 2024, Meatly announced it had reduced the cost of its chemically defined protein-free medium to an industry-leading £0.22/l, and in 2025, announced it had reduced the cost of bioreactors by ~10x. Following its regulatory authorisation in 2024, Meatly sold the world’s first cultivated pet food in 2025. Check out our earlier interview with Owen Ensor, founder and CEO of Meatly. This new funding will enable Meatly to build a 20,000-litre bioreactor facility in London, which is the largest of its kind in Europe. Fit-out of the facility will begin immediately, with product releases expected to follow in 2027.Investors in the round include Oyster Bay Venture Capital, Clean Growth Fund, and JamJar Investments.   This latest raise builds on the £7 million in seed funding provided by founding investor, Agronomics, and Pets at Home, bringing total funding raised to date to £17.4 million.   Owen Ensor, CEO, Meatly, sees the investment as a powerful endorsement, not just of Meatly, but of Britain's foodtech and biotech sectors.  “Meatly has one focus: to make commercially viable cultivated meat a reality. Over the last four years, Meatly’s pioneering team has systematically focused on reducing key costs and building the strongest possible technical foundation for growth. Now we have our own industry-leading technology, and we are ready to scale." Connor Duffy, Investment Manager at Clean Growth Fund, said:  "Rethinking how we produce protein is an essential part of tackling the climate crisis. We’ve invested in Meatly because they are showing it’s possible to produce real meat cost-competitively and with a fraction of the environmental impact.”   Elise Schumacher, Investor at Oyster Bay Venture Capital, said: “Meatly is not just building a new product - it’s laying the foundations for an entirely new protein category.      

As enterprises rapidly adopt agentic AI systems capable of autonomously executing tasks, interacting with tools, and making decisions across workflows, concerns around security, hallucinations, governance, and operational control are becoming a major barrier to deployment. French company Giskard is launching Giskard Guards, Europe's first independent sovereign guardrail platform for enterprise AI agents. I spoke to co-CEO and co-founder Jean-Marie John-Mathews to learn more. Enterprise AI agents face a growing range of risks, from hallucinations and unreliable outputs to security vulnerabilities, unsafe behaviour, regulatory compliance issues, and failures in how agents interact with tools, sensitive data, and real-world operational systems. Why AI agents remain vulnerable to manipulation So why does all of this occur? Well, simply put, AI agents are built on LLMs, and those models undergo training and reinforcement learning to make them helpful. In other words, training creates compliance, compliance creates vulnerability. “The challenge is that the model wants to remain helpful even when the request is malicious. That creates a major trade-off in how the systems are trained and fine-tuned,” explained John-Mathews. Further, many existing AI guardrails are ill-suited to the rise of chatbots and AI agents, as they were originally designed for social media moderation. “Traditional guardrails were mostly built for content moderation on social networks,” shared John-Mathews.  “With chatbots and AI agents, you’re not just moderating text — you also need to understand the actions the AI is performing.” A seemingly harmless request could trigger sensitive backend actions.  “A user might say, ‘Please delete the Q2 budget draft,’ but what matters is the underlying function call,” he shared. “Maybe the user doesn’t have admin privileges. Traditional moderation systems can’t handle that level of operational context.” Why conventional solutions fail  According to John-Mathews, there are currently two primary solutions to reduce these risks. You can do it offline—meaning large scans where you try to find many vulnerabilities in your agent. Or you can do it online, meaning guardrails that block unsafe responses. But existing generic LLM guardrails weren’t built for agents. They lack operational and company-specific context, and they can generate large numbers of false positives, even blocking up to 40 per cent of legitimate requests. Take a banking example. A user writes: “Fraud claim — what should I do?” A generic guardrail might block that because it sees the word “fraud”. But in context, it’s perfectly legitimate. The user may simply be asking what to do if their card has been compromised. “Without understanding context, you can’t moderate correctly,” explained John-Mathews. Further, traditional solutions are often designed around toy benchmarks (like blocking the forgetting of previous instructions) and fail to account for real-world attacks such as multi-step social engineering, context manipulation, or toolchain exploitation. The most common attacks aren't always the most dramatic. Often, says John-Mathews, they begin with someone simply trying to make an agent step outside its intended role — and succeeding. “There are things that are really dangerous, and things that are more reputational.” He offered the example, exposed by Som_patel5 on X, in which the poster opened the Chipotle restaurant chatbot to order food and explained that he needed help with a task: specifically, writing a Python script to reverse a linked list. In doing so, a supposedly limited-purpose enterprise AI assistant still exposed the broader capabilities of its underlying large language model, suggesting weak containment and inadequate guardrails.  While this not only leads to reputational damage, it also suggests that users may be able to steer the system outside its intended scope through prompt manipulation.  This could create opportunities for risks such as prompt injection attacks, extraction of hidden system instructions, misuse of connected tools or APIs, and potential access to sensitive backend workflows.  In another incident, the Chevrolet dealership chatbot became a viral example of how poorly constrained AI systems can be manipulated through prompt injection. Users discovered that the ChatGPT-powered bot could be tricked into ignoring its intended role as a car sales assistant and instead follow arbitrary instructions, including agreeing to sell a $76,000 Chevy Tahoe for $1. From AI testing to real-time agent protection Giskard’s Guards blocks unsafe answers and behaviours. Guards inspect the full agent execution chain (tool calls, parameter validation, multi-step reasoning) and adapts to each agent's business domain.  To achieve this, Giskard builds multiple detectors. Some are extremely fast, almost instant. Others are more contextual and deeper. Each detector maps to a specific action. If the system detects unsafe behaviour, it can block the response, notify developers, or simply monitor and log the interaction. Even before today’s launch, the company's platform for testing enterprise AI agents was used by Mistral, Google DeepMind, BNP Paribas, AXA, Doctolib, among others. The tech is suitable for deployment for sensitive agentic workflows in regulated sectors like banking and insurance. According to John-Mathews: “Customers kept pushing us past detection. They asked for runtime AI agent protection adapted to their context. Guards closes the loop: detect, then protect and fix vulnerabilities and hallucinations in real time.” The company also works extensively with healthcare organisations, where AI chatbot failures can quickly become high-risk. “Their chatbots can sometimes diagnose issues if users provide symptoms,” explained John-Mathews. “And depending on how the question is framed, the responses can become dangerous.” Further, conversational AI systems may reinforce harmful beliefs. For example, if a user expresses scepticism about antibiotics, the chatbot can become demagogic and suggest avoiding them. The same thing can happen with vaccines. The problem can also extend to manufacturing environments. “Imagine someone asking about electrical voltage requirements for a washing machine,” shared John-Mathews.   “If the chatbot gives the wrong answer, that could be a disaster.” From policy documents to enforceable code John-Mathews contends that traditional AI governance processes are struggling to keep pace with the rapid deployment cycles of enterprise AI agents.  While compliance teams still rely on manual risk assessments, Word documents, Excel spreadsheets, and annual audit cycles, AI teams are deploying and updating agents weekly — creating what the company describes as a dangerous governance gap that can lead to incidents.  Giskard's response is what John-Mathews calls "policy-as-code" — the idea that governance rules shouldn't live in documents at all, but should be embedded directly into the AI system as machine-enforceable logic. This approach replaces static documentation with machine-enforceable policies, automated compliance checks, Git-based version control, and rapid rollback capabilities, alongside prebuilt policy packs aligned with frameworks such as the EU AI Act and OWASP. “Whether the rules come from regulators or from internal company guidelines, we make them enforceable through the AI system itself. The system can be hosted directly by the company on European infrastructure and used to enforce safe behaviour.” Europe's sovereign AI stack is moving from policy to product The company also sees the rise of agentic AI as part of a wider push toward sovereign European AI infrastructure. A critical component of Giskard Guards is its position as a genuinely sovereign European AI security stack.  While “EU sovereign AI” has often remained largely political rhetoric, Giskard is attempting to deliver a concrete security-layer product: an independent European provider with its own technology, optimised models, and fully on-premise deployment capabilities. John-Mathews asserts:  “We want to turn European AI governance into actual products and enforceable systems — not just abstract policies.” For regulated sectors such as banking, insurance, and healthcare, sovereignty, data control, and deployability are no longer theoretical concerns, but operational requirements.”

In 2025, Europe’s healthtech ecosystem showed strong funding momentum, with capital concentrated in large rounds across biotech, medtech, digital health, and AI-enabled healthcare. The biggest deals came from companies such as Oura Health, Isomorphic Labs, Verdiva Bio, Tubulis, and Neko Health, highlighting investor interest in both consumer health platforms and advanced life sciences. By country, the UK stood out as the most active market, with major rounds across AI drug discovery, biotech, surgical robotics, digital care, and diagnostics. Switzerland also showed strong depth, particularly in biotech and medtech, while Germany, the Netherlands, France, Spain, and the Nordics remained important contributors. By round type, the market was led by a mix of large growth rounds and strong Series A and Series B activity. Series A deals showed continued appetite for early clinical and platform development, while Series B and later rounds reflected investor support for companies moving toward commercialisation, clinical expansion, or international growth. Overall, 2025 pointed to a maturing European healthtech market, where capital flowed to companies with clear clinical, technological, or commercial validation, while early-stage innovation remained active across a broad range of European hubs (for more detailed analyses of the European technology ecosystem, check out Tech.eu’s annual report: European Tech 2025 - The Big Picture). Here are ten healthtech companies that raised the most in 2025. Amount raised in 2025: $900M Oura Health is a Finnish company that develops the Oura Ring, a smart wearable that continuously tracks sleep, activity, heart rate, and other biometric data. Through its mobile app, it delivers personalised insights to support users’ health and wellbeing. In 2025, the company raised more than $900 million in a funding round that valued it at around $11 billion. The capital is intended to enhance AI capabilities, drive product development, expand global reach, and introduce new health features. Amount raised in 2025: $600M Isomorphic Labs is a London-based artificial intelligence company focused on transforming drug discovery through advanced machine learning. A spin-out of DeepMind, it builds AI models based on technologies such as AlphaFold to predict biological structures and design new medicines more quickly and efficiently, with the aim of accelerating breakthroughs in human health. In 2025, Isomorphic Labs raised $600 million in its first external funding round, led by Thrive Capital with participation from Alphabet and Google Ventures, highlighting continued backing from its parent ecosystem. Amount raised in 2025: $411M Verdiva Bio is a clinical-stage biopharmaceutical company developing next-generation therapies for obesity and cardiometabolic diseases. Its pipeline includes oral and injectable treatments designed to improve efficacy, convenience, and long-term outcomes, with a focus on patient-friendly options such as once-weekly dosing. The company leverages advances in gut-brain biology to create innovative medicines aimed at addressing significant unmet medical needs in global health. In 2025, Verdiva Bio launched with $411 million raised in a Series A for advanced obesity therapies. Amount raised in 2025: €308M Tubulis is a biotechnology company developing next-generation antibody-drug conjugates (ADCs) for cancer treatment. Its proprietary platform technologies enable the creation of highly stable, targeted therapies designed to deliver anti-cancer agents directly to tumours, improving efficacy while reducing side effects. The company is advancing a pipeline of ADC candidates for solid and hematologic tumours, with a focus on expanding the therapeutic potential of precision oncology and addressing areas of high unmet medical need. Tubulis raised €308 million in a Series C funding round in 2025 to advance the clinical development of its lead cancer therapy TUB-040, including expanding into earlier treatment lines and additional indications. Amount raised in 2025: $260M Neko Health is a healthtech company focused on preventive healthcare and early disease detection. It develops advanced scanning systems that combine sensors, imaging, and AI to collect and analyse large amounts of health data in a non-invasive way. Through its health centres, the company provides comprehensive body scans and immediate, doctor-led consultations, aiming to shift healthcare from reactive treatment to proactive prevention. In 2025, Neko Health raised $260 million in a Series B for preventative health scanning tech. Amount raised in 2025: €200M Ortivity is a Munich-based healthtech company building an integrated network of outpatient orthopaedic practices and surgical centres across Germany. Founded by physicians, it connects clinics, physiotherapy centres, and medical providers into a unified platform delivering diagnostics, treatment, surgery, and aftercare. Its model focuses on improving access, efficiency, and quality of care through a physician-led approach and the use of digital solutions, aiming to modernise orthopaedic services and shift care from inpatient to outpatient settings. Ortivity secured €200 million in 2025 to expand outpatient orthopaedic care. Amount raised in 2025: $200M CMR Surgical is a medical device company developing robotic systems for minimally invasive surgery. Its flagship product, the Versius surgical robot, is designed to assist surgeons in performing precise procedures while improving access to keyhole surgery. The company focuses on enhancing surgical outcomes, reducing recovery times, and making advanced robotic-assisted surgery more accessible across healthcare systems. In 2025, CMR Surgical raised over $200 million to expand global deployment of its Versius system, including entry into the US market, and to advance product development and innovation. Amount raised in 2025: €132M Azafaros is a clinical-stage biotechnology company developing disease-modifying therapies for rare genetic and metabolic disorders. Founded in 2018, the company focuses on small-molecule treatments targeting lysosomal storage diseases. Its lead candidate, nizubaglustat, is an orally available therapy designed to address underlying disease mechanisms, with the aim of delivering new treatment options for patients with limited or no existing therapies. In 2025, Azafaros completed an oversubscribed €132 million Series B funding round. Amount raised in 2025: $150M Cera is a London-based healthtech company providing digital-first home healthcare services. It combines a network of carers and nurses with AI-driven technology to deliver care, monitoring, and support in patients’ homes, aiming to improve outcomes, reduce hospitalisations, and shift healthcare from hospitals to community settings. In 2025, Cera raised a $150 million funding round to expand its AI platform, increase service diversity, and support operational growth. Amount raised in 2025: $150M Distalmotion is a medtech company developing robotic systems for minimally invasive surgery. Founded as a spin-off from EPFL, it created the DEXTER surgical robot, designed to simplify operations and integrate into existing hospital workflows. The company aims to expand access to robotic surgery by reducing complexity and enabling more surgeons and healthcare facilities to adopt advanced, precise surgical techniques across a range of procedures. Distalmotion raised $150 million in a 2025 Series G round to expand deployment of its DEXTER surgical robot in the US.

A UK university spinout, which builds full-stack quantum computers made with the same silicon chip technology used in laptops and smartphones, has raised $160m in an investment round, it said today. The Series C funding round in Quantum Motion was co-led by US VC DCVC and Spanish deeptech investor Kembara with participation from British Business Bank and Firgun. Also investing were Oxford Science Enterprises, Inkef and Bosch Ventures. Quantum Motion, which has raised over $200m to date, said it’s now the UK’s best-funded quantum computing company. The startup says that using silicon chips means it can build quantum computers more cheaply and more energy-efficiently than rivals. Leveraging silicon could mean a 100-fold reduction in cost and space requirements and a 1,000-fold reduction in energy consumption, the 2017-founded startup said. A full-stack quantum computer includes all layers required to perform quantum computing, including a Quantum Processing Unit (QPU), a user interface, and a control stack compatible with standard quantum computing software.  The funds will be used to commercialise its offering, as well as further R&D and geographical expansion, it said. The startup was founded by professor John Morton, based at the London Centre for Nanotechnology at UCL, and professor Simon Benjamin of Oxford University. The London-headquartered company also has offices and labs in Spain, Australia and the US. Doctor James Palles-Dimmock, CEO of Quantum Motion, said: “Today’s announcement reflects the strength of the team we have built and the progress they have delivered. Quantum computing will only achieve its full potential if it can be built on a platform that scales, and we believe silicon is the strongest route to achieving that."

Invest Europe, the association representing Europe’s private equity, venture capital and infrastructure sectors, has released Investing in Europe: Private Equity Activity 2025, its annual report examining performance across the region. The report provides a comprehensive breakdown of activity across segments, stages, sectors and geographies. It also introduces new data on continuation funds and investment in strategic areas such as defence and deep tech, supported by strong momentum in biotech and life sciences. According to the report, fundraising remained concentrated among fewer, larger funds, with pension funds and international investors playing an increasingly prominent role. Buyouts continued to drive activity, while venture investment showed signs of recovery, exceeding its longer-term average. Across segments, ICT remained the leading sector, alongside continued momentum in biotech and life sciences. The report’s findings highlight a resilient private capital market, with both fundraising and investment reaching their second-highest levels on record. Private equity and venture capital firms raised €147 billion in 2025, a 16 per cent increase on 2024 levels and second only to 2022’s record. Buyout funds were the primary driver, accounting for €103 billion, a 33 per cent increase, reflecting renewed demand from global investors despite ongoing geopolitical and economic uncertainty. Source: Investing in Europe: Private Equity Activity 2025 Total investment rose 3 per cent to €135 billion, marking the second-strongest year after 2021. Buyout investment reached €90 billion, broadly in line with 2024 and 16 per cent above the five-year average. The mid-market segment accounted for 34 per cent of total buyout value, reflecting continued capital allocation to SMEs and scaling businesses. Venture capital investment increased to €20 billion, 20 per cent above the five-year average. Source: Investing in Europe: Private Equity Activity 2025 The report also finds that activity accelerated in the second half of the year as markets stabilised following earlier uncertainty, with investors refocusing on long-term fundamentals. Exit value remained solid at €45 billion in 2025, supporting distributions to limited partners. Commenting on the findings, Eric de Montgolfier, CEO of Invest Europe, said: Private equity and venture capital activity showed great strength in 2025 against a backdrop of geopolitical and macro uncertainty. Europe provides a predictable and stable environment for global LPs to deploy their money, as well as being full of dynamic and high-potential businesses for skilled managers to back and grow. Overall, the report provides a detailed breakdown of activity across sectors, stages and geographies, and for the first time includes data on continuation funds and strategic investment areas such as defence and deep tech. 

Sweden Ballistics (SWEBAL), the Swedish defence manufacturing company driving domestic trinitrotoluene (TNT) production to strengthen NATO resilience, today announces a €30 million funding round to complete construction of its TNT manufacturing facility in Nora, Sweden.  SWEBAL is dedicated to strengthening NATO's resilience and enhancing European security through sustainable, large-scale ammunition production.  Founded in 2024, SWEBAL is committed to addressing supply chain challenges, including the limited production of critical materials like TNT. Having secured environmental permits from Sweden's Land and Environmental Court in December 2025 and approval of the detailed development plan in January 2026, SWEBAL is advancing construction plans for its new TNT facility, with the regulatory process proceeding as expected.  The funding will be used to complete the factory and unlock European supply chains at scale, at a time when foreign import dependencies and long-distance shipping pose a serious risk to the continent’s manufacturing.  The €30 million strategic investment is from several investors. Firstly, Sweden’s ex-Chief of Army, Major General (ret) Karl Engelbrektson.  In 2023, Karl took on the position of Head of Advisory Board at Finserve Global Security Fund (GSF). Karl’s personal investment in SWEBAL is independent of the GSF fund. Pär Svärdson, founder of Apotea, Sweden’s largest online pharmacy, and Adlibris, Sweden's largest online bookshop, is also an investor. Thomas von Koch, founding member and ex-CEO of private equity firm EQT. Alongside other prominent Swedish family offices.  According to Joakim Sjöblom, Co‑Founder and CEO of SWEBAL, the investment comes at a historic moment for European defence.  “Across the continent, governments are rapidly expanding military production capacity in response to ongoing geopolitical tensions, record defence spending, and persistent ammunition shortages. Europe has faced a critical shortage of energetic materials since Russia’s invasion of Ukraine, constraining ammunition output.  SWEBAL’s facility is set to produce over 4,000 tonnes of TNT annually, supplying the explosives European manufacturers need to sustain rapidly expanding artillery shell production, drone munitions, mines and other crucial armaments.”  Once operational, the facility will run 24/7 using European raw materials and machinery. Full-scale production is targeted for 2028 and is now fully financed.  Lead image: Joakim Sjöblom, Co‑Founder and CEO of SWEBAL. 

Ghent-based Tekst has closed an €11.5 million Series A led by US venture firm Elephant.  Tekst was founded in Ghent in 2022 by Wouter Janssen (CEO) and Tiebe Parmentier (CTO) and is taking aim at one of the biggest headaches in enterprise AI. Companies are pouring money into AI but struggling to make it pay off at scale. The problem, Tekst argues, isn't the technology itself — it's everything that comes before it. AI can read the data, but it doesn't grasp the context employees rely on to make decisions.  Where traditional tools optimise pieces of a process, Tekst goes after the whole thing — the messy, end-to-end work that runs the back office: quotes, orders, claims, customer service.   According to Wouter Janssen, CEO and co-founder of Tekst, the processes that actually run a business aren't written down anywhere.  “They live in emails, in PDFs, in people's heads. Our process intelligence technology surfaces all of it. Without that, an AI agent is just an expensive assistant or a glorified chatbot. The companies that crack this open get to use AI on a completely different level — and finally see real returns on what they've put in."     Rather than hiring consultants to spend months interviewing and analysing, Tekst reconstructs processes automatically from the digital trail employees leave behind every day. The system reveals which messages trigger which actions.    For most enterprises, these critical processes are often the least automated. At a typical Tekst customer, dozens or even hundreds of employees handle quotes and orders by hand: long email threads, order details buried in PDFs, contract clauses reviewed line by line.   Tekst’s platform connects directly to systems companies already use — SAP, Salesforce, Microsoft — and lifts the repetitive work off the table. Customers include Daikin Europe N.V., Colruyt Group, Securex, and Becton Dickinson.     The investment will fuel product development and international expansion. Tekst has 35 employees today and plans to double headcount by year-end, with hiring focused on engineering and go-to-market.   "The AI market is moving at breakneck speed, but Tekst has unique technology that maps and automates the most complex business processes in record time. What consulting firms typically deliver in months — with large teams and hefty costs — Tekst delivers in just a few weeks,” shared Pieterjan Bouten, General Partner at Entourage.      According to Matt Tanenblatt, Partner at Elephant, many mid-market and enterprise companies want to deploy AI agents — but they're missing the process intelligence layer needed to do it reliably.  “Tekst understands that gap and has developed an AI-powered product to help close it.  Their technology shows how processes actually flow within an organisation, and we believe that's the foundation AI agents need to operate properly in the back office."  Lead image: Tiebe Parmentier (left) and Wouter Janssen (right). 

Sedivention, a Germany-based medtech startup, has raised €2.9 million in a seed funding round led by bmp Ventures alongside the IBG funds. Additional investors include the strategic investment arm of a global medtech company, High-Tech Gründerfonds (HTGF), superangels, and Cambridge Ventures. The company is developing a minimally invasive, one-time outpatient therapy for the treatment of obesity, a condition expected to affect more than one billion people globally in the coming years. Existing treatment options, including bariatric surgery and drug therapies, remain limited due to their invasiveness, cost, or long-term dependency. Sedivention’s approach is based on a targeted cryo procedure delivered via a specially designed balloon catheter, similar to a gastroscopy. The therapy uses precise cryoablation to interrupt hunger-related signals in the vagus nerve, aiming to address the underlying physiological mechanisms of obesity and enable sustained appetite reduction without surgery, implants, or ongoing medication. Dr Ute Nollert, founder and Chief Medical Officer of Sedivention, said that obesity is a chronic disease requiring treatments that are both clinically effective and minimally invasive: Our approach targets the underlying disruption in hunger and satiety regulation, enabling a lasting reduction in hunger without invasive procedures or lifelong treatment. A functional prototype has already been developed and tested. With the new funding, Sedivention plans to advance product development, generate initial clinical data through a first-in-human study, and prepare for subsequent regulatory and market entry steps. In the long term, Sedivention aims to replace highly invasive surgical procedures with outpatient interventional treatments, improving access to effective care.

Stockholm-based Pit, an AI-native platform that replaces the patchwork of spreadsheets, inboxes, and rigid SaaS tools used in enterprise operations, has announced its public launch. The company also raised $16 million in funding led by Andreessen Horowitz (a16z), with participation from Lakestar, its founders, and angel investors including executives from OpenAI, Anthropic, Google, Deel, and Revolut, as well as the Stena and Lundin families. Pit positions itself as an “AI product team as a service,” enabling enterprises to build and run custom software tailored to their internal operations. The platform aims to replace the fragmented mix of spreadsheets, inboxes, and rigid SaaS tools that continue to underpin many business processes despite significant investment in digital transformation. Founded by former leaders from Voi, Klarna, and iZettle, Pit is designed to translate business needs into fully deployed, production-grade software. Its offering consists of two core components: Pit Studio, which learns how organisations work and builds tailored systems, and Pit Cloud, which provides governed infrastructure with enterprise-grade security and compliance features. Unlike traditional low-code platforms or AI copilots, Pit delivers fully operational software systems rather than prototypes. The platform is already being deployed in enterprise environments across sectors including logistics, telecom, e-commerce, and healthcare, with customers such as Voi, Tre, Stena Recycling, and Kry. For decades, enterprises have relied on software that dictates how they operate. AI now enables companies to run on systems designed around their own workflows, said Adam Jafer, CEO and co-founder of Pit. With the new funding, Pit plans to scale its platform and expand adoption among large enterprises seeking more flexible, AI-driven approaches to managing core business operations.

OpsMill, a Paris-based infrastructure data management company, has raised $14 million in a Series A funding round led by IRIS, with participation from BGV and existing investors Serena and Partech. OpsMill addresses a core challenge in enterprise IT: fragmented and unreliable infrastructure data. Its flagship product, Infrahub, is an open-source infrastructure data management platform built on a graph database with native version control. Designed as a system of record for IT environments, it enables infrastructure and network teams to scale automation and adopt AI-driven operations with greater reliability. As enterprises increasingly turn to automation, the quality of underlying data has become critical. Many organisations still rely on disconnected systems such as spreadsheets, legacy databases, and custom scripts, which can lead to errors and operational risks in automated or AI-driven workflows. OpsMill aims to provide a unified and trusted data layer to support safe, scalable automation across complex infrastructure environments. Damien Garros, co-founder and CEO of OpsMill, said automation fundamentally depends on data, and without a complete view of the network, organisations lack full visibility: We built Infrahub so that infrastructure teams and the AI agents working alongside them, always have a complete, trusted record of what exists and a way to safely evolve at scale. Infrahub models infrastructure as a network of relationships rather than static assets, allowing teams to manage dependencies, validate changes, and maintain governance through a structured DevOps workflow. The platform is available as both a free open-source Community edition and a commercial Enterprise version. The company has already seen adoption from large infrastructure operators and enterprises across sectors including retail, manufacturing, and financial services. With the new funding, OpsMill plans to further scale its platform and meet growing demand for AI-ready infrastructure data solutions.

London-based CodeWords, an AI automation platform, has raised $9 million in a seed round led by Visionaries, with participation from firstminute capital, Sequel, and Illusian. The round also included a group of angel investors and industry leaders such as Andrey Khusid (CEO of Miro), Mati Staniszewski (CEO of ElevenLabs), Hanno Renner (CEO of Personio), Robert Gentz (CEO of Zalando), Ilkka Paananen (CEO of Supercell), Alexandre Berriche (Founder of Fleet), Kieran Flanangan (CMO at Hubspot), and François Chollet (Co-founder of the ARC Prize), along with leaders at OpenAI, Mistral, n8n, and Zapier. CodeWords is built around Cody, an AI agent designed to learn how a business operates and proactively build and run automations across tools and workflows. Rather than requiring manual input or technical setup, the platform enables non-technical users to automate tasks such as deal flow monitoring, content creation, and lead generation across multiple integrations. The system operates entirely on CodeWords’ infrastructure, handling deployment, maintenance, and execution. The company originally started as an AI research lab under the name Agemo, focusing on neurosymbolic reasoning. Following early research recognition, the founders pivoted toward building a practical product aimed at simplifying automation for everyday business users. CodeWords has since moved into beta and continues to expand its capabilities, including contextual memory, messaging integrations, and adaptive execution modes. The best operators don’t wait to be asked. We built Cody on the same principle - an agent that learns about your business, sees what needs doing, and delivers outcomes, said Aymeric Zhuo, co-founder of CodeWords. With the new funding, CodeWords plans to scale its platform and further develop its AI agent, aiming to make automation more accessible and enable businesses without dedicated technical resources to operate more efficiently.

A device ships but collects field data at only 60% of the target rate. The hardware vendor blames firmware. The firmware contractor faults the RF environment. The RF consultant cites power management. No one modeled how modem power use, battery discharge, and packet retransmission interact as a system because no single team owned all of them. This ownership gap stalls IoT projects. This issue frequently delays and inflates budgets for many European device projects, especially for companies new to connected products or established manufacturers adding connectivity without a full redesign. ACRIOS Systems, a Czech technology company known for its large-scale smart metering deployments, built its custom development practice specifically to offer end-to-end solutions. Instead of fragmented service delivery, ACRIOS provides clients with full project ownership, handling every stage of the product lifecycle from embedded hardware design through OEM production to long-term field maintenance. Clients gain a single accountable partner who streamlines management and minimises integration risk for reliable results. The fragmentation problem in IoT product development Connected device creation spans circuit design, PCB layout, firmware, radio protocols, antenna tuning, power management, application software, backend systems, and cybersecurity. Each is specialised. Commonly, hardware, firmware, and cloud layers are split among separate teams. This may work initially, but integration often reveals gaps at these boundaries, with no single supplier accountable for resolving them. These gaps cause delays and rework. Hardware, firmware, antenna, or protocol issues often arise from split development. Single-team ownership avoids predictable failures. Full-stack ownership across the product lifecycle ACRIOS provides in-house hardware, firmware, embedded software, protocols, applications, backend services, and OEM production, all managed by a 25-person team without outsourcing core engineering. ACRIOS can manage end-to-end development or join an existing team for specific phases, such as optimisation, audit, or certification, without disrupting existing work. Prototypes are delivered within 2-3 months thanks to reusable tools, past architectures, and minimal coordination overhead when one team owns the full context. Communication expertise as a foundation IoT communication is the bedrock of ACRIOS’s engineering practice. Clients benefit directly. Robust device communication is critical for real-world performance, while mistakes in this domain have immediate, visible consequences. ACRIOS’s proven strength here is a decisive differentiator. The protocol stack includes bare-metal implementations of CoAP, MQTT, LWM2M, and UDP. It encompasses wM-Bus bidirectional communication and NB-IoT integration. A multi-chipset standardisation layer gives clients flexibility in hardware selection. FUOTA enables reliable, large-scale, over-the-air updates to deployed devices. Protocol validation occurs on active networks in several countries, not only in laboratories. Clients thus receive products tested in real-world environments, protecting their deployments from unexpected field issues. Radio behaviour in dense cities differs from that in the lab. Products not validated in real networks carry risks into production. ACRIOS is an official Quectel design house for Central and Eastern Europe. This status shows deep engineering skill in Quectel module integration and direct access to manufacturer support. For NB-IoT or smart module products, this relationship shortens integration challenges. It removes long back-and-forth with component vendors. Designing for demanding environments Industrial, utility, and public infrastructure deployments face demanding conditions that typical consumer IoT development overlooks. With portfolio highlights such as a converter certified to ATEX Zone 2 for explosive environments, ACRIOS ensures devices meet stringent requirements for enclosure design, power management, and component selection. The company’s hands-on experience translates to robust, field-ready solutions. ATEX-certified hardware experience directly informs every aspect of device designs. Power budgets are conservatively calculated. Environmental tolerances are guided by field conditions, not lab scenarios. Mechanical design accounts for real-world technician use, not just engineers in controlled settings. This approach guarantees reliability where others falter. Discipline applies beyond utilities. For Lokni, a smart water vending machine, ACRIOS delivered control systems, a mobile app for users, an admin web interface, and a secure backend. The device is maintenance-free, using remote diagnostics for long uptime and reliable management. Cross-sector experience strengthens solutions. "A device that works in a lab is a prototype. After three years in the field, two firmware updates, and a radio regulation change, it is a product. That distinction shapes every architectural decision we make from day one," says Marek Novák, Chief Technology Officer at ACRIOS Systems. Cybersecurity compliance as an engineering discipline Europe's regulatory environment for connected devices is changing fast. The Cyber Resilience Act took effect in 2024, with compliance due by 2027. It enforces cybersecurity for the full lifecycle: secure development, vulnerability management, encrypted communication, OTA updates, and certification documentation. CRA compliance is not a late-stage add-on. It requires strategically planned architectural choices from the start. This means robust authentication, sound encryption key management, and prepared responses to post-deployment vulnerabilities. Retrofitting these features later is costly and inefficient. With a development team prepared for these challenges from the start, compliance becomes simple and cost-effective. ACRIOS takes part in applied research funded by the Technology Agency of the Czech Republic. This research focuses on practical CRA requirements to help clients benefit from firsthand regulatory expertise. The company designs device architectures and prepares certification documentation, ensuring clients receive products that are not only operational but also ready for European certification. The Radio Equipment Directive (RED) and its delegated acts add cybersecurity and privacy rules for radio devices. ACRIOS aligns both CRA and RED requirements from the architecture stage. This ensures compliance through design, not last-minute fixes. From embedded device to data layer For clients needing more than devices, ACRIOS manages network connectivity, backend infrastructure (Dockerised Linux), and data export through various interfaces (REST API, CSV via SFTP, etc.). The practical consequence for product teams is the ability to work directly with business-relevant data rather than managing an IoT infrastructure. The device collects. The backend processes. The client's systems receive clean, structured output, reducing internal technical workload. For companies new to connected products, this approach eliminates the burden of building and maintaining their own IoT backend, delivering a turnkey solution with a lower barrier to entry and reduced operational overhead. The same architecture supports remote monitoring and OTA updates, making long-term deployments viable and CRA-compliant. Devices can be updated, rolled back, and managed remotely as requirements evolve. The case for single-partner development A single engineering partner across the full stack ensures clear accountability, fast fixes, and swift regulatory adaptation when requirements change. For European companies bringing connected products to market under increasing regulatory pressure and with limited tolerance for extended development cycles, that accountability structure is a material advantage. The alternative, distributing development across multiple suppliers and managing the integration, is not cheaper. It transfers integration risk to the client, who is typically least equipped to absorb it. ACRIOS has validated this model across metering infrastructure deployments, public environment hardware, and hospital communication systems. The technical domains differ. The underlying engineering discipline does not. p.p1 {margin: 0.0px 0.0px 5.0px 0.0px; font: 12.0px 'Times New Roman'; color: #000000} About ACRIOS Systems ACRIOS Systems is a Czech technology company specialising in hardware and software development for smart metering, IoT, and energy management. An official Quectel design house for Central and Eastern Europe, the company designs and manufactures its own hardware and firmware in-house, and delivers full-stack custom development and OEM production for clients across Europe. For more information, visit ACRIOS Systems.

An AI consulting-cum-recruitment startup co-founded by a Google DeepMind scientist and a SoftBank executive has raised $22.75m, led by new investor Andreessen Horowitz.   The Series A founding round in Ethos also features participation from General Catalyst, Matt Miller's Evantic and trading firm XTX. It follows its $3.25m seed round last year led by General Catalyst. The startup has raised around $30m in total. Ethos, based in London, leverages AI to help foundational AI labs, investment funds and corporates with recruiting experts in specific fields, for consulting and full-time roles. As well as talking to a candidate via voice-powered AI, the startup's AI scrutinises a candidate's portfolio of work, including academic papers, Github repositories, and professional content, matching them up with the expertise that its corporate clients want.   It says that more than 5,000 candidates are joining each week across accounting, banking, consulting, law, technology and healthcare alongside skilled tradespeople, including electricians and plumbers.   Ethos was founded by CTO Daniel Mankowitz, who spent six years at Google DeepMind, and James Lo, a former McKinsey and SoftBank Vision Fund executive, who is now the CEO of Ethos.   The funding will be used to develop Ethos’s AI agent and expand its global network, while scaling partnerships with AI labs and enterprise clients, it said. Lo said: "A CV is a poor proxy for what someone is truly capable of. Ethos helps people to see the full shape of their expertise and where it fits in an AI-transformed economy - acting as their agent, opening doors they didn't know existed, and giving companies a far more precise way to find the people they actually need.    “AI is reshaping the labour market faster than our tools for valuing human expertise can keep up. Ethos is built to change that.”

OpenTrade, a stablecoin yield infrastructure platform used by fintechs and exchanges, has raised $17 million in a strategic funding round led by Mercury Fund and Notion Capital, with participation from a16z Crypto, AlbionVC, and CMCC Global. The round brings the company’s total funding to over $30 million. The raise comes as the global stablecoin market surpasses $300 billion in supply, driving demand for infrastructure that enables secure and scalable yield generation. OpenTrade provides plug-and-play solutions that allow fintechs, neobanks, and exchanges to offer dollar- and euro-denominated yield products backed by real-world assets, without building their own investment or custody systems. Since its launch, the platform has gained traction with partners such as Littio, Midas Kripto, and Glim. It has surpassed $200 million in total value locked and processed more than $250 million in transaction volume in 2025, reflecting growing adoption across both retail and institutional use cases. As demand has evolved, OpenTrade has expanded beyond its core infrastructure to include a permissionless protocol layer and Curation+, a suite of vault curation services designed for more complex, institutional-grade strategies across real-world and on-chain assets. These capabilities enable asset issuers, non-custodial platforms, and treasuries to access diversified yield strategies without managing underlying operational complexity. David Sutter, co-founder and CEO of OpenTrade, said the company has simplified how fintechs and neobanks integrate institutional-grade stablecoin yield into their offerings: As we grew, it became clear that our infrastructure could also serve non-custodial platforms, treasuries, and asset issuers that all need the same thing: a safe, scalable way to connect stablecoins to diversified yield strategies. The company has also deployed its infrastructure as a permissionless protocol that issues transferable, position-tracking tokens, with its first live implementation through Sierra Protocol. In parallel, Curation+ formalises its investment strategy services, combining regulated asset management oversight with active portfolio design and execution. The newly raised capital will be used to expand OpenTrade’s infrastructure, further develop its Curation+ services, and grow its engineering, asset management, and customer success teams as it scales globally.

In modern warfare, speed is becoming the defining advantage. Systems that once took years to design and deploy are now expected to reach the battlefield in months — or risk becoming irrelevant before they are ever used. RSI Europe is a Lithuania-based defence tech company building remote-controlled explosive systems and drone technologies designed for rapid deployment in active conflict environments.  I spoke with CEO Tomas Milašauskas on a visit to Vilnius to learn how the company achieved rapid time-to-market and profitability from year one. From energy regulation to defence tech Tomas Milašauskas started his career as a public servant in energy regulation, setting electricity prices in Lithuania and working on European-level infrastructure projects, including interconnectors with neighbouring countries. After that, he moved into consulting, and then became a fund manager in renewable energy — investing across Eastern Europe in solar, wind, energy storage, and other EU-aligned projects. Image: Tomas Milašauskas, CEO of RSI Europe. Culture as a competitive edge in a company born out of urgency “In Lithuania, defence isn’t just about the army — it’s a whole-of-society effort,” says Milašauskas. “It includes the military, volunteer organisations, and industry.” That ethos is reflected internally at RSI Europe. “A large part of our team has direct links to defence — former or current military personnel, volunteers, or people like myself involved in organisations such as the Riflemen’s Union,” he explains, referring to the civilian force of around 17,000–18,000 members trained to support national defence. This shared context shapes how the company operates. “There’s a clear understanding of the threat, and a very clear goal,” Milašauskas says. “When you’re building these systems, you know they could ultimately protect your own country, your family, your loved ones. That responsibility gives the work real meaning — and it’s what keeps us pushing forward.” ​ That context became reality overnight. Russia’s full-scale invasion of Ukraine changed everything. When everything escalated, the company's owners had a moment of: "What do we do now?" “It reached a point where it felt like a direct threat to Europe,” he explained. “That’s when my brother and I decided to launch a company focused on defence technology — specifically to help Ukraine." Neither of them came from the armed forces, so they relied on what they did have — their finances and technical thinking. They started with remote initiation systems — RISE-1 systems — something they understood better because of their background in technical communications. Speed: from concept to deployable product Within a few months, as a team of just three people, they had already developed rough 3D-printed prototypes being tested by the Lithuanian Armed Forces. After the first round of feedback, the second iteration was already deployed in Ukraine. Initially, the team approached the military with ideas around smart, remotely controlled munitions — essentially IoT-enabled systems. “They told us: that sounds great, but your time to market will be at least five years,” recalled ​​Milašauskas. If the goal was to help Ukraine win, they needed something deployable within 3 to 5 months. So RSI Europe pivoted to remote initiation systems — focusing on the “smart” part without the munition itself. That allowed it to get into the field quickly. RSI Europe stands out for its speed from idea to time-to-market. Milašauskas contends that if your development cycle is two or three years, your only option is to sell to NATO. “If you want to sell to Ukraine, it has to be two to three months — maximum. The battlefield evolves constantly, so you have to react quickly.” This is also about a shift in mindset: ”It’s not about having a perfect, polished system. It’s about whether it works. For example, a drone costing around €1,000 can destroy a tank worth several million. So the question becomes: do you want a perfect product, or an effective one?” Starting with a niche product According to Milašauskas, RSI Europe saw two paths — drones and remote initiation systems. “We chose the smaller, less competitive product. In 2022, there was no defence funding available. Competing in drones would have been unrealistic for a small team with no capital. So we focused on a system that had no real competitors. We started in Explosive Ordinance Disposal and expanded into broader military use — including special operations and intelligence units.” Today, those systems are widely used for unexploded ordnance (UXO) clearance — especially in heavily mined areas following Ukrainian counteroffensives — allowing operators to initiate explosives remotely while staying at a safe distance, sometimes up to 25 kilometres away. ​ Its technology stack extends to LoRa-based radio communications, autonomous and semi-autonomous drone capabilities, and FPV drones optimised for precision strike operations in high-electronic-warfare conditions. Image: An RSI Europe Shpak FPV drones. How RSI Europe’s technology works Technically, the system architecture is relatively simple. At its core, it’s an IoT device. There is some edge processing — for example, drones receive simple commands and translate them into real-time control across multiple axes. But the compute itself is relatively low. What really differentiates military systems is the networking layer: radio links must be highly resilient and capable of withstanding jamming and interference, using techniques such as advanced modulation and limited autonomy when signal conditions degrade. In addition, these systems operate at significantly higher power levels than consumer devices, ensuring more reliable connectivity in contested environments. “As we move towards more autonomous systems, compute requirements increase — but for now, many systems remain relatively simple,” explained Milašauskas. ​Profitable from year one The company was initially bootstrapped with a personal loan of around €40,000 — a modest starting point by startup standards, particularly in a market where early-stage companies often raise significant capital before generating meaningful revenue. Yet within its first year of operation, the business generated approximately €170,000 in revenue, establishing a strong commercial foundation from the outset. ​ “We’ve been profitable since year one,” shared Milašauskas. ​ The funds provided the flexibility to focus on the product and the end user, rather than investor pressure. Scaling challenges: logistics and deployment A key challenge for the company going forward is logistics. Milašauskas explained that Ukrainian companies can deliver within 1 to 2 weeks. “For us, exporting adds another two to three weeks. We also don’t always know where the product will end up — there’s often a gap between delivery and deployment. That’s why we maintain local teams in Ukraine to support users and gather feedback.” ​ From hardware to doctrine Beyond logistics, the challenge shifts from technology to application. Where do drones sit within the military structure — infantry, engineering, artillery? To address this, the company has worked closely with the Lithuanian Armed Forces to develop a playbook for integrating these systems into real operations. As a result, its offering now goes beyond hardware and software to include practical operational know-how. ​ Autonomy and real-world limits Tomas Milašauskas notes that some level of autonomy is already in use, including features such as terminal guidance, position hold, return-to-home without GPS, and early forms of swarm coordination. However, development remains cautious. “In a chaotic environment, control is critical. If you lose control, you risk unintended consequences,” he says. He adds that much of what is shown publicly does not reflect operational reality. “Many public demonstrations take place in controlled environments — they’re not representative of real battlefield conditions.” Looking ahead, Tomas Milašauskas points to two priority areas: building explosives capabilities — where he sees a clear gap in Europe — and advancing autonomy in drone systems. ​ Fragmentation, variety, and consolidation Drones are fundamentally reshaping defence procurement, shifting it away from standardised, single-platform systems toward a model where diversity of capabilities — not uniformity — defines operational strength. Tomas Milašauskas argues that, unlike traditional defence systems, drones require diversity rather than standardisation. “With something like a fighter jet, you invest in one platform — it’s extremely expensive to maintain, requires specialised technicians, and takes years to deploy,” he says. With drones, variety is the capability itself. Different systems perform differently depending on the environment. A drone that works in one part of the battlefield may be completely ineffective in another. This shift is likely to reshape procurement models. “Users will increasingly adopt a mix of capabilities rather than relying on a single system. That could mean multiple suppliers — or one supplier offering a range of specialised drones for different use cases.” At the same time, the market is evolving rapidly. “It’s a massive space, with many players and a lot of capital coming in,” Milašauskas adds. “Some are consolidating by acquiring smaller companies, others are building full ecosystems and trying to become the next primes. And many will simply disappear. In the first year, five companies were supplying the Lithuanian military. Now it’s down to one or two — and new players continue to emerge.” Startups should focus on specialisation Tomas Milašauskas advises founders to focus on specialisation rather than building yet another drone. “There are already many large players, and even more coming out of Ukraine at scale, so competing head-on with a new drone platform is very difficult,” he says. Instead, he points to opportunities in the gaps around the ecosystem. “There’s strong demand for specialised components, particularly as non-Chinese supply chains are still being built — and not yet fully there. Even basic elements, like magnets, often still come from China.” He adds that innovation can go far beyond the drone itself. “There’s a lot to be done with modular electronics, new components, and edge AI. It’s a strong market, but you need to be focused. Identify what’s already crowded, find the niche that isn’t, and build there."

A Finnish AI startup, co-founded by an executive whose previous startup was sold to US semiconductor giant AMD, has raised €25m in an angel round, giving it a €325m valuation, just months after its launch. Helsinki-based Qutwo was co-founded by Peter Sarlin, whose previous startup Silo AI was sold to AMD for $665m in 2024. Qutwo has raised the round from angel investors, which include Legora’s Max Junestrand, Hugging Face’s Thomas Wolf and co-founders and owners from Schwarz Group, Index Ventures and Atomico, amongst others, it said. Qutwo’s ambition is to build Europe's leading AI lab for the quantum era. Qutwo is building a software platform called Qutwo OS that, it says, helps companies get more out of AI, as quantum computing develops. It says it has secured more than €20m in contracted revenue since its February launch. Qutwo was founded by Sarlin, Kaj-Mikael Björk, Sarlin’s co-founder at Silo AI, and Kuan Yen Tan, co-founder at Finnish quantum outfit IQM. It says it already employs more than 50 scientists and engineers from the likes of Microsoft Research, UC Berkeley, and and Cambridge University. Sarlin said: “Qutwo’s angel round provides us more than just capital. We are at a moment that comes once in a generation. “We’re honoured to be backed by founders and investors who have built and backed category-defining global companies, as we build the platform for the next AI paradigm."