According to the Tech.eu database, in July, European tech companies collectively secured €9.3 billion across 355 deals, an 18 per cent increase compared to June’s €6.5 billion across 323 deals.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 July Tech.eu Pulse today.

OpenAI today launched the latest version of the technology that underpins its popular ChatGPT bot, claiming GPT-5 was its “smartest, fastest” and “most useful” AI model to date. The arrival of GPT-5 has been highly anticipated and comes as ChatGPT hits nearly 700 million weekly active users.It also comes amid a heated battle between the likes of OpenAI, Google, xAI, Meta and Anthropic as they rush out new AI models, which they hope will trump each other on benchmark metrics. GPT-5 is targeted at retail and business customers, but a blog post announcing its launch focused on its use for businesses, championing it as introducing a "new era of work", as OpenAI looks to acquire enterprise clients. San Francisco-based OpenAI said GPT-5 offered users improvements from previous models across several areas, including reliability and accuracy and that it suffers from fewer hallucinations. It also said GPT-5 provided “more accurate” answers than its previous AI reasoning models and heralded its writing credentials and its ability to answer health-related questions.“GPT-5 delivers leaps in accuracy, speed, reasoning, context recognition, structured thinking, and problem-solving”, OpenAI said.OpenAI also bigged up GPT-5’s coding credentials, saying it was its “strongest coding model yet” as it looks to wrest market share from Anthropic, whose Claude model is popular with coders.It also alluded to its vibe-coding credentials, saying users without coding prowess can create “beautiful” and “functional” websites, as well as its ability to power AI agents. Other new features include advanced voice, so the bot can better understand instructions, and an option for free and paid users to colour their chats.The arrival of OpenAI's latest proprietary model follows its release of two “open-weight” models earlier this week, as it looked to take on Meta’s Llama and Chinese models.GPT-5 offers both free and paid options. A blog post announcing the arrival of GPT-5 came with a string of endorsements from firms including Uber, Salesforce and Moderna.

Deeptech wildlife intelligence company Flox has closed an oversubscribed Seed round of nearly $1 million. Spun out of  AI research at KTH (Royal Institute of Technology) in Stockholm, the company aims to make human-wildlife interaction safer, smarter, and more harmonious, using robotics and AI to enable safer, more sustainable coexistence between humans and wildlife.     Flox has developed Edge — a fully autonomous device that speaks to wildlife in real time using adaptive bioacoustics powered by AI. The autonomous, battery-powered Edge units patrol even the most remote areas within a perimeter with onboard AI, instantly detecting wildlife and deploying species-specific bioacoustics to steer animals away, without the need for fences or downtime. No fences, no downtime.   Connected to the Flox Wildlife Platform, it delivers 24/7 protection for both wildlife and critical infrastructure across agriculture, aquaculture, railways, and aviation.  The first production batches for 2025 are already fully booked, and Flox is already collaborating with partners like WWF, Alstom, Boliden Minerals, GRR Airport and Swedish Transport Administration.     “Flox is the first company that truly aims to understand and speak the language of wild animals — and that will change everything. With the explosion of AI capabilities, our deep understanding of wildlife behaviour and partnerships with leading wildlife research institutes and NGOs, the timing is perfect. We’re sitting on enormous untapped potential to put AI to good and redefine how industries interact with nature, not just for protection but for real coexistence,” said Sara Nozkova, CEO and Co-founder of Flox.     The round was led by Unconventional Ventures, with participation from Norrsken Accelerate, and all investors from Flox’s Pre-Seed round: Almi Invest, E14 Invest, Argand Partners, and a group of bold angel investors.    According to Nora Bavey, General Partner at Unconventional Ventures, Flox is tackling one of the world’s most overlooked yet urgent challenges: human-wildlife conflict through a highly scalable, tech-driven solution.  “With a bold vision, deep sector knowledge, and early traction across multiple industries and geographies, the team is poised to transform how we coexist with nature. We’re proud to support Flox on their mission to turn a global problem into a sustainable, cross-border opportunity.”      After four years of research and validation trials with wildlife biologists and leading enterprises, the Seed round will help fuel the commercialisation of the solution and continued R&D, with a clear focus on expansion in the United States.  These goals will be achieved through a combination of venture capital, non-dilutive research grants the company has secured, and growing revenue from commercial deployments, positioning Flox for rapid growth and global market entry. This round gives Flox the momentum to scale a whole new category of tech and redefine the boundaries of coexistence, starting with airports, transportation and industrial sites to keep both wildlife and people safe and keep operations going.

As AI reshapes everything from banking to entertainment, one German startup is betting it can do the same for fitness.  Zing Coach has built an AI-powered workout app that aims to be smart, adaptive, and personalised. The company, headquartered in Munich, has operational teams in Warsaw and Cyprus, and it raised $10 million in June 2024. It combines proprietary health data, real-time adaptation, and computer vision to deliver tailored exercise plans that evolve alongside each user's needs.  At the helm is new CEO Anton Marchanka, an entrepreneur with deep roots in the fitness tech space, who sees Zing as the beginning of a paradigm shift toward agentic AI in fitness. I spoke to Marchanka to learn more. Adaptive fitness, powered by real-time data Zing Coach offers a highly personalised, data-driven approach to exercise. It builds individualised training plans based on a user's unique body composition, fitness level, and workout history, drawing on data from its own Fitness Test and Body Composition Scan, as well as Apple Health.  Powered by AI, Zing continuously adapts each user's program in real time, adjusting intensity, complexity, and duration based on factors like fatigue, progress, and training environment, whether at home or in the gym. With over 400 guided exercises, Zing Coach uses computer vision technology to track users' movements, helping them improve form, avoid injury, and make measurable progress.  The platform customises workouts to suit available equipment, target muscle groups, and even health restrictions, positioning itself as a smart companion that evolves alongside its users' goals. Marchanka has extensive experience in mobile marketing and product management. At Mosaic Group Holdings, he served as General Manager and led a portfolio of 20 mobile apps with 600 million users globally. Previously, he was Vice President of Business Operations at Daily Burn as well as Vice President of Marketing at Apalon. His experience spans portfolio management, product strategy, and digital user acquisition. The next generation of health apps is personal by design Marchanka believes there are three key trends at play in health and fitness: AI-powered personalisation.: Everyone from small startups to companies like Whoop and Oura are using AI to create hyper-personalised experiences. Data as a foundation: AI without meaningful user data is useless. Companies with access to rich, proprietary health data will define the future of this space. Holistic health:  Fitness is no longer just about workouts. It's about sleep, stress, nutrition, and mental well-being. People are realising that to get real results, they need to address everything, not just training. He asserts that there's also a shift in focus.  "It's no longer about whether you lose 20 pounds by your wedding. It's about how you feel throughout the process. Enjoying the journey matters just as much — if not more — than reaching the goal." Further, unlike the early days of app stores, there's more attention given to the veracity of wellness apps, which often fall short due to a lack of scientific rigour, personalisation, and accountability. To avoid these pitfalls, Zing works with medical institutions in Switzerland, and all of its content is developed in collaboration with experts — especially when it comes to users with specific conditions like joint injuries. Marchanka asserts that AI has resulted in a shift from cohort-based personalisation to individual personalisation. "Pre-AI, we grouped users based on demographics or behaviour. Now, we can treat each user uniquely, just like a real doctor would. We consider your heart rate, sleep, mobility, preferences, and craft the experience around you. Our goal isn't to replace human coaches. It's to amplify their capabilities and make expert guidance scalable. AI agents should be extensions of real people, not substitutes." "A full circle moment" in mobile apps In terms of the state of the mobile app market, Marchanka believes we're seeing a "full-circle moment—almost like we're back in the early 2010s when the App Store was still new and app development felt revolutionary." "That shift moved people from web to mobile. It created billion-dollar companies like Duolingo that could only exist because of the app ecosystem." He asserts we're on the cusp of another paradigm shift: from apps to AI agents.  Zing is building an AI agent that can eventually outperform even our most experienced human experts.  "This isn't just buzz. We're moving away from rigid, use-case-based applications to flexible AI agents that can reason, learn, and adapt. It's like we've gone from building tools to training companions." "In five years, I believe everyone will have their own personalised AI that assists with everything from communication to decision-making. This changes the whole dynamic of how we build and use software," shared Marchanka. I was curious what this means for developers: are they becoming obsolete, or more empowered? Marchanka recalled that a former colleague once told him there are two kinds of developers: code writers and value creators. Code writers, who primarily translate specifications into syntax, are the most vulnerable to AI automation. But value developers, the ones thinking in terms of architecture, systems, and impact, are more essential than ever. AI has eliminated a lot of the boring, repetitive work in app development, and thus, developers who embrace it can focus on higher-order thinking and solving meaningful problems. "In my ideal team, we'd have just a few highly capable developers managing the system's architecture while the AI does the rest. So I see software development evolving into something more like a craft, or even an art. Less code, more creativity," he shared. Virality is no longer a strategy for app-building startups  For startups launching an app today, it's a brutally competitive landscape. Why do so many fail to gain traction? Marchanka asserts that the bar is much higher than in the easy days of the App Store when just getting the right keywords or app name could shoot you to the top of search results: "Now, you're effectively locked into paid marketing. Organic growth still exists, but it's hard. And even platforms like Meta or TikTok, which still allow for some creative virality, are becoming more commercialised." He sees TikTok as a game-changer: "It's still possible to reach large audiences by working with micro-influencers and creating highly engaging, short-form content.  But even that takes investment. You need to budget for creative, influencer coordination, and production. If you're starting from zero today, your best bet is to build a product with strong retention and engagement, then raise funds to amplify it. Trying to rely solely on virality is like buying a lottery ticket." Reflecting on the pace of technological change, Marchanka looks to his own family:  "I often think about my grandmother. She was born in a small village in 1930, where even having a telephone was rare. By the time she passed away in her mid-80s, she'd lived to see the age of AI agents. That's a massive shift in one lifetime. Now it's our turn to shape the next leap — and I want Zing to be at the forefront of that journey."

Bit2Me, the digital asset platform in the Spanish-speaking market, has secured a strategic investment from Tether, the issuer of USDT and a global leader in the digital asset space. This strategic acquisition positions Tether as the main backer of an investment round that will total €30 million over the coming months, alongside the participation of additional strategic investors. Bit2Me is Spain’s digital asset company, with over a decade of experience building crypto infrastructure and serving the Spanish-speaking market. Recognised for its reliability and award-winning exchange services, Bit2Me also drives crypto education through Bit2Me Academy, the largest Spanish-language platform for learning about cryptocurrencies. Andrei Manuel, co-founder and COO of Bit2Me, commented: Having a global leader like Tether as part of our shareholder structure elevates Bit2Me to a new level within the international crypto ecosystem. Together, we aim to lead the market’s development in both Europe and Latin America—a region that is beginning to embrace decentralised finance. Bit2Me has always followed a clear path grounded in legality, transparency, and a strong commitment to doing things right for our customers.” Pablo Casadio, co-founder and CFO of Bit2Me, noted that the company is supported by leading investors and institutions, including Telefónica, Unicaja, Cecabank, BBVA, and Investcorp. He added: With  over a decade of experience, we’ve earned the trust of more than 7,000 businesses and millions of users worldwide. Tether’s support marks a pivotal moment for Bit2Me to scale further in customers, operations, and global reach. Tether Ventures, based in El Salvador, is spearheading the operation as part of its global strategy to support impactful tech projects in both emerging and developed markets. Tether’s investment in Bit2Me, combined with Bit2Me’s newly obtained EU regulatory license and the backing of future strategic partners, will enable the Spanish company to accelerate its expansion across the EU, while also reinforcing its footprint in Argentina and other parts of Latin America, consolidating its position as a key player in both the European and Spanish speaking crypto ecosystems. Paolo Ardoino, CEO of Tether, stated: Bit2Me has shown a strong commitment to building secure, transparent, and regulation-compliant infrastructure with a user-first approach. Their focus on education, transparency, and usability aligns with Tether’s vision of an open financial system that empowers individuals through technology. We’re excited to support their growth and role in developing regulated crypto-asset services in Europe and beyond. Lead image: Bit2Me team | Photo: Uncredited

Environmental monitoring technology company WATR has secured a growth capital investment from 24Haymarket to accelerate its expansion and meet the growing international demand for real-time water and soil quality data. Founded in 2017, WATR provides an integrated solution for remote environmental monitoring, offering continuous real-time data on water and soil quality. Its flexible platform supports data-driven decision-making for clients across sectors such as river catchment management, agriculture, utilities, construction, fisheries, and environmental consultancy, helping them protect ecosystems and meet regulatory standards. WATR’s proprietary, cloud-connected platform is hardware-agnostic and compatible with third-party sensors. It supports multiple telemetry protocols, including GSM, LoRa, and Satellite, enabling reliable monitoring even in remote areas. Data is delivered through an intuitive dashboard and alert system, allowing users to respond swiftly and effectively. Glyn Cotton, CEO of WATR, shared: This investment gives us the fuel to reach the next stage of our mission: to improve water and soil quality worldwide. We’re focused on helping customers make better decisions, faster - by making high-quality environmental data more accessible and actionable than ever. This investment enables the business to offer the availability of ‘Flexible Ownership Models’: Outright purchase or Data-as-a-Service (DaaS) options. WATR’s technology is already deployed across the UK water sector and internationally in Germany, Portugal, Croatia, France, Saudi Arabia, Australia, and most recently, the United States. The business has experienced rapid growth over the past 12 months, driven by increasing global demand for scalable, accurate, and easy-to-deploy environmental monitoring solutions. Jamie Dunnett, Investment Director at 24Haymarket, said: WATR is delivering critical infrastructure and real-time data solutions at a time when environmental monitoring is a global priority. The team’s technical agility and deep sector understanding position them to be a leading player in this fast-growing space. The new investment will fuel WATR’s commercial growth by scaling its sales team, strengthening operations, and enhancing marketing to accelerate adoption in existing and new markets. It will also support the company’s strategic partnerships, product development, and international expansion. Additionally, WATR will continue to invest in innovation, advancing its real-time sensing platform with AI-driven insights, predictive analytics, and deeper integration into customer ecosystems.

Cybersecurity startup WiseBee has raised $2.5 million in pre-seed funding to accelerate the development and deployment of its Agentic cyber risk mitigation platform. WiseBee was founded by Stoyan Stoyanov (CEO) and Taha Kazi (CDO), who lead a globally distributed team that brings together deep cybersecurity experience and cutting-edge expertise in AI and machine learning. Their goal is to transform how security teams work through an AI-powered platform. WiseBee redefines the traditional cybersecurity approach by not only identifying threats but also automatically resolving them through AI-driven remediation workflows, delivering enterprise-level security to mid-sized organisations. The company is already working with lead customers, including a local US government, as well as top fintech, banking, and space tech institutions across Europe. Stoyan Stoyanov, CEO of WiseBee, commented: Most cybersecurity tools are reactive, siloed, and designed for large enterprises with hefty budgets. We are building something fundamentally different: an autonomous system of security agents that detect and mitigate risks around the clock - 24/7. Cybersecurity remains one of the most resource-intensive functions in modern organisations, often relying on complex tools and large teams to stay ahead of threats. For mid-sized security teams, this creates a critical gap with too many alerts and too few people to act.  WiseBee secures critical public and private sector environments, supporting fintechs, space tech, government, and regulated financial institutions with AI-driven risk management and automated threat resolution. Pilot programs are underway with CISOs in fintech, banking, government, defense, and telecom. By enabling real-time resolution without additional headcount, WiseBee cuts time-to-response from days to hours. The round was led by Frontline Ventures and BrightCap Ventures, with participation from a US-based investment firm and angel investors with cybersecurity experience. Frontline Ventures, which backs ambitious tech companies across the US and Europe and positions them to win the transatlantic market, recognised the urgency of WiseBee’s mission.  Zoe Chambers, Partner at Frontline Ventures, shared: AI brings huge benefits but also new security threats, from rogue agents to unvetted tools and chaotic tech stacks. Stoyan and Taha are building exactly what stretched infosec teams need: AI-native agents that go beyond alerts to deliver full remediation. We've been impressed by their speed and traction so far and are excited to back them as they scale across Europe and the US. Dimitar Korsakov, Principal at BrightCap, added: Cybersecurity is no longer just about better alerts; it’s about intelligent context and precise autonomous actions. Wisebee delivers on all fronts, thanks to Stoyan and Taha's deep industry expertise, AI innovations, and customer-first approach.We are delighted to partner with Wisebee as they bring their product to thousands more organisations and continue growing their team with top AI talent. The funding round is also supported by a network of strategic advisors, including CISOs and engineers from SecurityScorecard, VMWare, Box, Amazon, Deloitte, Verizon, and Google, who are veterans at building enterprise security, automation, and data tools. WiseBee is actively expanding across Europe and North America, with early traction in CEE, the UK, DACH, and the US West Coast. With this new funding, the company will strengthen its platform’s autonomous capabilities, broaden integrations, and grow its team to support product-led growth. WiseBee is also laying the foundation for its next funding round and promoting a community-driven approach to cyber defence. Lead image: Freepik

UK challenger bank Zempler Bank has reported its third consecutive year of pre-tax profit, as customer deposits and interest income increased. Zempler, which focuses on microbusinesses, reported pre-tax profits of £5.2m in the year ending 31 March 2025, a 58 per cent increase on the previous year.Zempler, which was founded in 2005 as Cashplus, reported revenues of £68.2m in the period. The challenger bank, which garnered a UK banking licence in 2021, focuses on microbusinesses which it says are underserved by high-street banks.Customer deposits grew 15 per cent to £626m while Zempler recorded a slight uptick in its number of business customers to just over 170,000 open accounts. Interest income was up from £34m to £37m over the year, while total income increased by more than seven per cent to £68.2m. At the end of 2024, Zempler, which rebranded to its new name last year, announced it would be laying off around five per cent of its staff. Zempler’s total headcount as of the end of March 2025 was 338.Rich Wagner, CEO, Zempler Bank:  "Zempler’s growing profitability will be a key driver of our strategy over the next 12 months as we continue to optimise our business, strengthen our capital position and diversify our business through lending growth. "The year ahead is set to hold challenges for many of our customers and we know that confidence in the UK economy and the spending power of consumers is under pressure from macroeconomic conditions and that many small firms are facing an increasing tax burden."

Tracelight, a UK-based AI company simplifying the creation and validation of financial models, has raised $3.6 million in seed funding to advance product development, grow its team, and expand its reach to finance and strategy professionals globally. Complex financial models form the backbone of global finance, guiding mergers, investments, and strategic decisions worth trillions. Yet this critical domain remains largely unaffected by the AI revolution. The tools used for financial modelling have seen little change in decades. Microsoft Excel, still central to most modelling workflows, marks its 40th anniversary this September. Tracelight bridges the gap between financial models and large language models (LLMs) by turning spreadsheet logic into LLM-friendly data, allowing financial analysts and consultants to harness the benefits of Generative AI when building complex financial models. By integrating directly into the spreadsheet workflow and enhancing how LLMs interpret and work with Excel, Tracelight eliminates repetitive financial modelling tasks. From writing complex formulas and validating models to autonomously running analysis from simple natural language prompts, it is a force multiplier for analysts, augmenting their skills with AI, letting them model faster and smarter without needing to change the way they work. Peter Fuller, co-founder and CEO of Tracelight, said:  Complex financial models underpin all of the corporate world's most important decisions. Until now, LLMs have barely touched this workflow. Tracelight is purpose-built for this; it works seamlessly inside the tools analysts already trust. Our aim is to take away the entire mechanical workload of financial modelling. Our users act as directors and orchestrators of AI that is quickly becoming superhuman at the task of Excel analysis. That vision is supported by a deep technical foundation.  Since its launch, Tracelight has attracted early adoption from analysts at investment banks, private credit, and private equity firms, and top professional services providers. Early users report over 90 per cent time savings on tedious modelling tasks such as creating standard analyses, formatting, and error detection. Crucially, Tracelight keeps humans in control of decision-making. Rather than replacing analysts, it enables them to build better financial models and focus on high-stakes decisions where human judgment is essential. Aleksander Misztal, co-founder and CTO of Tracelight, added: The financial models used by investors, consultants, bankers and corporate finance professionals are large, logically complex, and formatted according to strict rules. Our technology unlocks the full reasoning power of frontier LLMs on these files; for the first time, Excel financial modellers will be able to tap into the power of frontier AI models directly in their existing workflow. The round was led by Chalfen Ventures, with participation from Acequia Capital, Inovo and EF, and angels including Charlie Songhurst (ex Microsoft head of Corporate Strategy) and Suhit Gupta (ex-CIO General Atlantic and Carlyle). Mike Chalfen, Solo VC at Chalfen Ventures, said:  The most magical AI tools work at the highest level of abstraction, making them radically easy to adopt, while understanding and executing complex work under the hood. Tracelight does that. It has taken the world’s most widely used modelling platform and embedded AI where it can have immediate impact. And its rapid adoption by analysts in demanding modelling environments shows just how big the opportunity is. Hank Vigil, Founder of Acequia Capital, noted that much of the world’s business logic still relies on Excel, even as it reaches its 40th year. He adds: What excites me about Tracelight is how naturally it brings advanced AI into the financial modelling workflows professionals already know and trust. It’s the first solution that extends the power of Excel modelling without requiring professionals to abandon the tools they know best. Tracelight is now available as an Excel add-in through the Microsoft Add-In Store, initially targeting individual analysts, consultants, and finance professionals. The newly secured funding will support further product development, team expansion, and broader distribution to reach finance and strategy professionals worldwide. Lead image: Tracelight co-founders | Photo: Uncredited

The Ukrainian government today announced the launch of its first state infrastructure for AI development, the AI Factory, and efforts to build its own national LLM.. According to the Vice Prime Minister for Innovation, Education, Science and Technology Development — Minister of Digital Transformation of Ukraine, Mykhailo Fedorov, the AI Factory is part of a mission to enter the top 3 countries in terms of AI development and implementation by 2030.  One of the key factors for its implementation is the availability of infrastructure on which AI services will operate, such as computing clusters and data warehouses. Fedorov shared:  “Soon, Ukraine will have the most modern hardware and software in the world for training and operating our state AI services. Requests in state AI services will be processed faster, and most importantly, our data will remain within the country. This is critically important for the technological security of Ukraine.” The AI Factory encompasses:  Technological Infrastructure: High-performance computing clusters (GPUs), water-cooled server rooms, data storage systems, and all the essential hardware to fully power AI development. Software Ecosystem: Tools for training and deploying models, data preparation interfaces, monitoring systems, and automation frameworks. Data Capabilities: Integration with national registries, along with tools for data cleaning, annotation, and transfer — ensuring models have access to high-quality, reliable information. Expertise and Talent:  Training programs for technical specialists who will design and implement AI solutions across public sector and defence applications. The AI Factory will deliver core local services developed by the WINWIN AI Centre of Excellence under the Ministry of Digital Affairs. WINWIN is the Ukrainian Global Innovation Strategy until 2030, designed to empower governments and international partners to drive economic recovery, strengthen sovereignty, and win peace through breakthrough technologies, open markets, and global collaboration. Ukraine has AI-powered tools ready or about to launch, including an AI assistant on the Diia portal designed to streamline access to public services — a major leap for its digital state. The Ministry of Digital Transformation is also deploying AI tools for analysing regulatory acts and translating European legislation, integrating AI into the MRIIA platform, and developing internal AI tools. Its also incorporating AI into defence tech as well and drafting Ukraine’s National AI Strategy, which it aims to complete by year-end. The need for a sovereign LLM for Ukraine Ukraine is also creating its own LLM. A domestic LLM allows Ukraine to run services without storing data abroad or in insecure jurisdictions. It also offers great benefits to defence-tech startups.  According to  “Ukraine is collecting uniquely rich battlefield data — unlike any other nation — and can leverage an LLM to process it and generate battlefield insights. This is critical from a security perspective.” The national LLM will be trained on Ukrainian data: history, scientific research, library collections, and other public-sourced materials. Official data will also be included, excluding any sensitive information.  Crucially, this will all be done in accordance with IP laws. Authors will also have the option to donate their works to enrich the model with authentic Ukrainian context. The LLM will be available to public institutions and businesses to create products and services, advancing both the public and private sectors. To achieve this, the Ukrainian government is partnering with Kyivstar, the largest Ukrainian mobile operator. The structured collaboration includes: A Coordination Committee to oversee vision and strategy. A Technical Board managing architecture, data, and model training. An Ethics Board ensures rigorous oversight on data usage and ethical considerations. Kyivstar, as the operational lead, will form a project office, recruit the team, provide computing resources for preliminary model training, and finance the entire LLM development. No state budget funds are being used. Post-launch, the model will be accessible to government bodies, research and educational institutions, and civic organisations. After a testing phase, it will be open-sourced — available for businesses and developers to use and build upon. According to Fedorvo, “Together, our goal is to build a high‑quality, scalable model within approximately nine months.” Building and maintaining its own AI infrastructure is a cornerstone of Ukraine’s digital sovereignty. It enables the secure development and deployment of AI solutions for citizens, the Defence Forces, and the public sector. Lead image: Mykhailo Fedorov, Vice Prime Minister for Innovation, Education, Science and Technology Development — Minister of Digital Transformation of Ukraine. Photo: uncredited.

In Europe, many of us are privileged to live in walkable cities with an abundance of micromobility, public transport, and rideshare options. In many places, cars are simply not built for the stresses of modern cities. From traffic to parking, owning a four-wheeler in urban environments often proves more of a hindrance than a boon. However, it's an uncomfortable truth that SUV sales are increasing.  SUVs dominate — even where they don't belong. In 2021, SUVs accounted for approximately 48 per cent of passenger car sales in China. In Europe, buyers also shifted away from hatchbacks and sedans, with SUVs — albeit including hybrid and EVs — making up 45.5 per cent of new car purchases, compared to 52.3 per cent in the United States. By 2024, the SUV trend had only intensified: SUVs represented roughly 54 per cent of all new passenger vehicle registrations across major markets in Europe. However, vehicles are smaller than a lot of the SUVs in the US.  According to one startup, it doesn't have to be this way, especially as most car use involves one or two people travelling a short distance and carrying a relatively small amount of stuff in something constructed from over 1800kg of metal.  The car sits idle for over 90 per cent of the day, taking up space. A lighter vision for Europe's roads Luvly, a Swedish startup based in Stockholm, believes there's a better way — and CEO Håkan Lutz says the future of European roads is light urban electric vehicles. For nearly 20 years, Lutz has been at the forefront of the mini-mobility movement as a developer and pioneer in reimagining urban transportation. A passionate entrepreneur and business professional, he surrounded himself with like-minded individuals who share his interests in cars, the environment, climate, safety, and impact. Luvly's core offering is a patented flat‑pack platform that enables the design, shipping, and rapid assembly of lightweight electric vehicles called Light Urban Vehicles (LUVs). Using sandwich-composite chassis panels and energy-absorbing foam modules, their architecture achieves high safety, low weight (between 380 – 400 kg), and — critically — shipping efficiency. A single container can fit as many as 20 unassembled vehicles, massively reducing CO₂ emissions and logistics costs. The vehicle uses very small motors and two removable battery packs — just one-tenth the capacity of a typical EV — which can be easily charged like an e-bike, eliminating the need for charging stations or battery-swapping infrastructure. Further, while it could be technically considered a microcar, Luvly O is the same size as a first-generation smart car, comes loaded with safety tech and has a trunk space of up to 269 litres and easily fits Lutz, who stands at 193cm (6' 4). A platform for others to build on Rather than compete head-on with legacy OEMs, Luvly licenses its platform to automakers and mobility operators, allowing licensees to build their own LUV models using Luvly's chassis and manufacturing process. Luvly also provides a full-support toolkit covering concept development, engineering, prototyping, testing, and production ramp-up. Why many microcar startups fail — and why Luvly hasn't EV's, especially Light EVs and microcars, are a tough environment for startups.  Over the last few years, we've seen the failure of Uniti, Carver, and even more abroad, such as Arcimoto, which I've had the pleasure to ride in over the Golden Gate Bridge in San Francisco.  Lutz suggests that there are two key reasons that Luvly has been able to succeed:   "First, we went deep on the tech. We're not using the same technology as everyone else—we developed a truly different solution, and that lets us do things that are incredibly hard to combine: sustainability, affordability, and safety in a single product." Second, people underestimate the difficulty of bringing a new vehicle to production and marketing it globally.  "It's incredibly costly and full of caveats. There are dominant players in this industry for a reason.  We realised, over time, that the right way forward is to offer something to the experts — companies that already know production, marketing, and how to navigate market constraints — and give them a solution to do something new, something they don't typically do." Significantly, Luvly fits into existing automotive OEM supply chains.  Everything used in a Luvly is essentially the same as what you'd find in a traditional car, just smaller and lighter.  Further, Luvly manages to create a safe but still light EV, with a safety cell that can be protected, with very affordable materials, while remaining strong and light. Lutz admits, "I won't claim it's as safe as a tank — but I also don't think we should be driving tanks on city streets."  "Because the whole thing weighs less, we can downsize components. It's like: every nut you remove reduces weight, and that lets you remove another one elsewhere." Challenging the SUV mindset When it comes to the increasing dominance of SUVs, Lutz asserts that while Luvly can't single-handedly change the entire market — there are about 2 billion cars out there —  it believes there's a market that hasn't yet realised what it needs. "People don't always know what they want until they see it. And there are beautiful small cars out there, but very few that could claim to truly be a replacement for a car," asserts Lutz.  The company ensures it's possible to offer consumers an alternative: something that does the same job as a car but uses a fraction of the resources. According to Lutz, Europe is a natural fit for vehicles like Luvly – an affluent population, a strong desire to reduce climate impact.  But he shared that to the company's surprise, they've had massive interest from the US too.  "And in Asia, ultralight vehicles are already the norm — tuk-tuks, rickshaws, etc.— but safety is often lacking. We think we can help there." When it comes to the auto industry's response to lightweight, low-cost EVs, Lutz believes that the industry is somewhat stuck: "Big cars, with more equipment, deliver higher margins. Simpler, smaller vehicles have low profitability, which is a problem. But what consumers — and the planet — really need are smaller, simpler vehicles. We hope to enable that shift by providing a way to build safe, sustainable, profitable ultralight cars." Fortunately, interest in microcars persists. The LUV and microcar movement is evolving slowly but steadily Image: Kate. French company Kate acquired the legacy brand Nosmoke and is developing the Kate K1, a more affordable, four-seater microcar aimed at everyday use, targeting a 90 km/h top speed, ~200 km range, powered by lithium batteries and a €15,000 price point. France is also home to Circle Mobility and Squad Mobility, although these companies are yet to deliver a vehicle to market.  In terms of conventional vehicle OEMs, Microlino sells two types of vehicles, and Honda recently launched the N-One, its smallest four-wheel vehicle.   And in terms of manufacturing innovation, In the UK, Helixx is developing a 'Factory in a Box' solution that enables customers to build mini commercial electric vehicles (EVs) anywhere in the world. Further, microcars and LEVs are part of car-sharing fleets such as drivemycar in Switzerland. A united front for policy change Luvly is also a member of The Microcar Coalition, a European alliance of electric vehicle startups promoting L7e-class "fast microcars" as a sustainable and space-efficient alternative to traditional cars. The coalition advocates for regulatory support, equitable subsidies, and urban planning policies that prioritise compact electric vehicles over large, high-emission SUVs. The coalition's goal is to reshape urban mobility by elevating microcars as a practical solution to reduce emissions, reclaim public space, and support more inclusive, efficient transport systems across Europe. While the Luvly O prototype has not yet entered certified road production, Stellantis is currently evaluating Luvly's flat-pack vehicle tech for possible integration with Stellantis' urban EV portfolio.  Lutz contends, "everything we've seen from them suggests they're serious—this isn't just a token project."

Green search engine Ecosia is now delivering its own search results for the first time, marking a major shift in the European tech landscape.  This milestone is made possible by the company’s joint venture with Qwant: European Search Perspective (EUSP). EUSP has developed Staan (Search Trusted API Access Network), a search index aimed to develop sovereign, privacy-first search infrastructure for Europe.  Check out our recent interview with Christian Kroll, CEO of Ecosia. For the first time in its 16-year history, Ecosia users in France will now receive a proportion of their search results directly from EUSP’s independent European index. The rollout aims to serve 30% of French search queries by the end of the year.  “This is a milestone moment for Ecosia and European tech,” said Christian Kroll, CEO of Ecosia “Having our own search infrastructure is a critical step for digital plurality and for building a sovereign European alternative. With more control over our offering, we can better serve users, develop ethical AI, and double down on our mission to build tech that benefits people and the planet.”  Much of Europe’s search, cloud, and AI layers are built on American Big Tech stacks, putting entire sectors — from journalism to climate tech — at the mercy of political or commercial agendas. Building a fully independent search index is more than a symbolic step — it's a foundational investment in Europe’s long-term competitiveness and democratic control.  A step toward autonomy, competition and sustainability Ecosia and Qwant have historically relied on syndication platforms, but by jointly building back-end technology and a European index, both firms are gaining greater autonomy over their tech —- and creating a viable, competitive alternative for other search engines and AI providers in Europe. “This isn’t just about better search - it’s about the freedom to build and shape the future of tech in Europe,” added Kroll.  “Regulators and lawmakers increasingly recognise the importance of competition and sovereignty in digital infrastructure. We hope they’ll continue supporting open, privacy-first innovation.” Open accessible European tech for everyone Unlike Ecosia’s existing steward-owned model, EUSP is structured to allow outside investment, enabling long-term scaling of its infrastructure. EUSP’s search index is also available to other tech companies, offering a foundation for competition, data privacy, and innovation in areas like generative AI. 

At a time when investors are falling over themselves to invest in AI, Marta Gaia Zanchi, Founding Partner at Nina Capital, takes a critical approach.  She contends, “We’re not anti-AI. We’ve backed many AI-based companies. We just care about whether it’s the right solution for the problem. That's what matters most.” Nina Capital is a Barcelona-based healthcare technology-focused investment firm founded in 2019, making on average ten investments per year in Pre-seed and Seed startups. The Firm announced the first close of its €50M Fund III in February this year. The AI flood: “Half the companies that pitch us say they use AI” According to Zanchi, the ubiquity of AI in the startup ecosystem is hard to ignore.  “The bulk of emails we get are either written by AI or about an AI-based company. In preparing for this call, I looked at our database — we've screened around 9,500 healthcare tech companies since we started in 2019. We’ve built internal tools, some of which are ironically using AI, to query that data. What we’ve seen is that nearly half — about 4,000 or so — of those companies claim that AI is a major component of their product. So we’re definitely somewhere near the peak of a bubble when it comes to AI, even if it's not quite the top yet.” She believes that many companies overstate their AI impact, noting, “Often what they’re describing as AI is just standard software — explicit, scenario-specific logic. Old-fashioned code, basically. It feels like they’re catering to investor appetite for AI, rather than honestly describing the tech. And I would actually place some of the blame on our own asset class — for fueling that bubble by rewarding the AI label more than the underlying value.” Many founders inflate their use of AI — and investors aren’t helping. Zanchi believes startups are misled about how investors perceive AI: “We’re not excited by the word 'AI.’ We’re excited by founders who truly understand the healthcare landscape and can speak to the perspectives of all its stakeholders. If they can show us why their solution fits and why it’s likely to succeed, that’s what matters.” Inside Nina Capital’s value-based approach to health innovation Nina Capital follows a Stanford Biodesign–inspired, need‑driven, value‑based methodology, in which innovation begins by deeply understanding real unmet needs in healthcare, then designing technology and business models to meet them, with rigour. This approach has been shown to de-risk venture creation in health technology since its inception in 2001. Nina Capital divides healthcare investment into three categories:  Clinical indications: Software or hardware/software combinations intended for clinical use. These are often regulated medical devices.  Infrastructure for healthcare businesses: Tech solutions that help pharma, biotech, medical device manufacturers, and pharmacy networks operate more efficiently, improve revenue, cut costs, etc. Care services – This includes: Tech-enabled care providers (virtual or hybrid care models that weren’t possible before). Tech products sold to existing care providers (like hospitals) to help them operate better. Health data tools — products that process, aggregate, or anonymise healthcare data to ensure privacy and improve interoperability. According to Zanchi, when it comes to hype, the most concerning examples tend to be in the first category, where companies are claiming to use AI for diagnostic or therapeutic purposes.  > “That’s very close to the point of care, and it introduces real risks if the tech isn’t robust.” Creating solutions in search of problems instead of understanding real value Zanchi also sees another issue. Even among companies that are building AI models, many are creating solutions in search of problems. Instead of starting with a healthcare need and asking what the best solution is, they’re pushing AI into spaces where it might not be the right fit. Healthcare is complex. Products need to appeal to multiple stakeholders, including patients, doctors, nurses, administrators, insurers, and regulators.   anchi contends that the first step is understanding the value proposition for each of those groups. What does this solution offer them? Is there alignment? “We often reach conviction — or decide to pass — before we even open the black box of the tech. If it turns out to be AI under the hood, great. But we’re looking for the best product to solve the problem, not the flashiest tool,” she shared.  Nina Capital backs diverse founders making tech that solves real healthcare problems In terms of what makes a good startup to invest in, in a word, diversity.  Zanchi explained:  “At Nina Capital, we’re a mix: 12 nationalities across 10 people, with backgrounds in engineering, neuroscience, pharma, medtech, regulatory, and SaaS. We look for that same diversity in our founding teams.” While early-stage teams can’t have everything in-house, an ideal starting founder triangle is someone who understands the clinical context, someone who brings technical expertise, and someone who has business or market knowledge.  Further, she contends that sometimes healthcare insight comes from lived experience, such as founders who built a product because of a personal or family medical experience.  "That kind of commitment can drive real innovation.” Examples of  investment include: Noah Labs: A digital health startup that develops Ark, an AI-powered, Class IIa telemonitoring platform combining smart biosensors, machine learning, and a mobile app to detect and predict heart failure decompensation — often up to 14 days before clinical deterioration—for earlier intervention and reduced hospitalisations  CryoCloud: a cloud-native SaaS platform that automates cryo‑electron microscopy (cryo‑EM) data analysis using machine learning, dramatically speeding up 3D protein structure visualisation to accelerate drug discovery. LillianCare: Establishing hybrid general practice clinics—where nurses handle about 60 per cent of outpatient treatment under remote tele‑supervision by doctors, enabled by an integrated digital platform and partnerships with insurers and municipalities “Growth alone won’t save you.” Zanchi views AI with a historical lens. “Think of the dot-com bubble. The internet didn’t fail — clearly — but there was a moment of overexcitement and overinvestment, and then a collapse." She believes AI will follow a similar trajectory. “It’s here to stay and will become a foundational technology. But we do need to be cautious.” Within this, she points to two recurring issues in these hype cycles: Jumping into tech before defining the problem. Prioritising growth over profitability. “The best companies in our portfolio can flex: they can grow quickly when the market rewards it, or shift to profitability when needed. But others are still suffering from the excesses of recent years,” shared Zanchi. In terms of adoption in healthtech, a sector with rightful caution and long procurement cycles, Zanchi contends that the key to access for startups is understanding the financial incentives that drive adoption and behaviour. Nina Capital’s most successful portfolio company founders know not only how patients flow through the system, but how money flows. They also understand what behavioural changes are required and what blockers exist. “They’ve been able to pivot when needed — if they hit a dead end with a stakeholder group, they adjust their positioning. They reconfigure the product to fit the workflows of healthcare providers better and redistribute the value so all stakeholders win.” “You’re not just promising 'AI,' you’re answering the tough questions: How will this impact patient outcomes? What do regulators think? Will administrators adopt it? Can it be reimbursed? Is privacy protected? When you can confidently answer those questions, then it doesn’t matter what’s in the tech stack—whether it's traditional software or deep learning. It’s about solving the problem." And they’re humble enough to recognise where their expertise is lacking and bring in advisors or experts to help. The most successful startups innovate with the system, not against it. Startups ignore the unsexy problems at their peril In terms of innovation gaps, Zanchi admits that some of the most impactful problems are also the most boring, citing an example of a conversation where an executive at a large US primary care group shared his biggest headache: “He said: vendor management. He’s trying to track technology spend, measure impact, and identify redundancies across departments — all with spreadsheets. There’s so much inefficiency there. But because it’s not “exciting” or directly patient-facing, it gets ignored. If you could move the needle on that, it would have a massive financial and operational impact, and ultimately benefit patients, too.” Further, Zanchi contends that many founders overlook how healthcare reimbursement and incentive models vary between countries: “Germany, France, the US — they're all very different. And those incentives change over time, sometimes rapidly. Founders need to be more attuned to that. When it works, it can be game-changing. Successful relationships with hospitals can last for years." Lead image: Nina Capital.

Deutsche WertpapierService Bank (dwpbank) has acquired Berlin-based fintech lemon.markets, taking full ownership to expand its full-custody services by offering a Brokerage-as-a-Service (BaaS) solution. In the future, the dwpbank Group will operate under two brands, combining many years of securities expertise and stability, with innovative strength. Established financial service providers and fintechs will benefit from a flexible range of services that allow them to adapt swiftly and effectively to market shifts, helping them attract and retain new customer segments in the securities business over the long term. With this move, the dwpbank Group (consisting of dwpbank and its subsidiaries) is advancing its growth strategy to build an ecosystem of modern, accessible, and reliable securities services tailored to the diverse needs of the financial sector. Partnering with lemon.markets enhances the Group’s cloud-based technology stack, adding streamlined, digital brokerage and custody infrastructure to its existing WP3 securities platform and wpNex digital asset solution. lemon.markets, a Berlin-based fintech founded in 2020, provides banks, fintechs, and asset managers with easy, digital access to capital markets. Its API-based platform integrates seamlessly into existing systems and supports features such as fractional trading, real-time processing, and fully digital workflows for ETFs, stocks, and funds. As a BaFin-licensed securities institution, lemon.markets serves prominent fintech clients, including Tomorrow, Optio, and Holvi. Kristina Lindenbaum, Board Member for Client and Digital Transformation at dwpbank, says: With the technological know-how and innovative offering from lemon.markets, we can now offer Banks, as well as Fintechs and Asset Managers, the best of both worlds in the securities business - expertise and stability combined with innovative strength. Within the dwpbank Group, we are enabling our partners to capture growth opportunities in a securities market increasingly shaped by neo- and online brokers. The dwpbank Group will pursue a two-brand strategy in the market, with both companies maintaining their independent identities while working closely together to leverage their respective strengths. Their collaboration will focus on jointly advancing their platforms, ensuring rapid innovation cycles, and offering a flexible service portfolio that enables established financial institutions and fintechs to adapt quickly and effectively to market changes. Through the integration of both platforms and joint development efforts, clients and partners will gain access to a broader, more competitive range of services, from full custody to digital Brokerage-as-a-Service solutions, that can be combined and tailored to their specific customer needs. Max Linden, Founder and CEO of lemon.markets, says: Both companies share the mission to make investing in capital markets more accessible - the best foundation for our cooperation. Alongside our technological leadership, we’re sending a signal of strength to existing and future partners and customers. With accelerated time-to-market, a modular service offering, and a modern customer experience, together with dwpbank, we are making financial service providers more competitive. Through this partnership, both companies are actively shaping the evolving securities market. By building a comprehensive ecosystem for securities services, they are showing how customer focus, technological innovation, and regulatory expertise can come together to establish new standards in the industry. The lemon.markets management team will continue in their current roles. The acquisition remains subject to approval by the German Federal Financial Supervisory Authority (BaFin) and customary closing conditions, with completion expected by autumn 2025.

In 2024, the Greek tech sector demonstrated remarkable momentum, expanding by 15 per cent, one of the highest growth rates among European ecosystems. Notably, artificial intelligence emerged as the dominant investment theme, attracting around one-third of total capital raised, a clear signal of where both local innovation and global investor interest are heading. A key driver behind the country's progress has been strong institutional support. Programs like Elevate Greece, alongside various EU-backed initiatives, have played a vital role in expanding the country’s venture capital infrastructure. Founders also benefit from favourable regulatory conditions, including simplified company formation, R&D tax incentives, and Golden Visa options for foreign investors and skilled talent. Together, these factors have firmly positioned Greece as a rising tech hub, marked by growing access to capital, dynamic entrepreneurship, and sectoral strength in emerging technologies like AI. These Greek tech companies are worth watching in 2025 reflecting the country’s expanding innovation and entrepreneurial strength. Amount raised in 2024: $45M Blueground is a proptech company offering fully furnished, move-in-ready apartments for medium- to long-term stays in 100+ cities. Founded in 2013 in Athens and headquartered in New York, it blends hospitality with technology to provide flexible leases, 24/7 support and a seamless digital experience. Serving professionals, students, and remote workers, Blueground redefines modern living with 40,000+ curated homes and a mission to help people feel at home wherever they go. In 2024, Blueground secured $45 million in Series D funding, reaching a bit over $1 billion valuation. Amount raised in 2024: €30M Hellas Direct is a tech-driven, digital-first insurance company founded in 2012, aiming to disrupt Greece’s insurance market. Offering fast, affordable, and flexible online insurance for cars, motorcycles, homes, trailers, and finance via Wallet+, they emphasise instant quotes, transparent pricing, and 24/7 customer support. With innovations like daily insurance, “FastTrack” claims, and partnerships with 3,000+ garages, they serve over 1 million customers across Greece, Cyprus, Romania, and beyond, committed to sustainable operations and safer roads. In 2024, the company secured a €30 million investment aiming to strengthen its climate-focused insurance offerings. Amount raised in 2024: $16M Harbor Lab is an Athens-based maritime technology pioneer founded in 2020 by Antonis Malaxianakis. It offers an AI-powered, cloud-native platform that streamlines port call management and Disbursement Account (DA) analysis for shipowners, charterers, agents, and vendors. With capabilities like instant DA cost estimation, transparent expense validation, secure payments, and automated claims, Harbor Lab delivers efficiency at scale, supporting 2,000+ vessels, 5,000+ service providers, and 35,000+ port calls. Backed by a $16 million Series A in 2024, its AI tools like HL AI enhance tariff analysis, language translation, and negotiation workflows to cut administrative overhead and empower smarter, data-driven decisions. Amount raised in 2024: €8.6M Brite Solar is a nanomaterials and solar-tech innovator developing semi-transparent photovoltaic solar glass and electrochromic smart glass for agriculture and construction. Their agrivoltaic panels enable greenhouses and open-field crops to generate clean energy, protect plants, save water, and boost yields, often achieving 10× more output while using 1/10th the water compared to traditional greenhouses. Brite Solar launched the world’s first semi-transparent solar panel manufacturing line, combining nanotech with scalable production. The company’s mission is to advance sustainable agriculture, reduce buildings’ energy footprint, and accelerate net-zero energy solutions globally. In 2024, the company raised €8.6 million for improving solar glass technology. Amount raised in 2024: €2M Wikifarmer is a global agri-tech platform. It empowers farmers worldwide by combining a free, expert-curated knowledge library of over 15,000 articles in 17 languages with an AI-enhanced B2B marketplace that directly connects producers and buyers, eliminating intermediaries, improving fairness, and boosting farmer profitability by up to 300 per cent. Through its Marketplace, Library, and Academy, Wikifarmer offers scalable tools, training, and market access to thousands of producers, transforming global agri-food supply chains. Wikifarmer secured €2 million in 2024 to boost its international expansion and growth. Amount raised in 2024: €1M Finloup is a fintech that offers businesses flexible, eco-friendly leasing of top-tier tech—smartphones, laptops, tablets, with low monthly fees. By promoting a circular economy and reducing e-waste, it enables clients, from freelancers to enterprises, to access the latest devices without upfront costs. Models include pay-as-you-go or fixed 12/24 month plans with built-in repair coverage and easy upgrades and ownership options. Finloup also pioneered BNPL via open banking, aiming to foster financial inclusion and sustainability. In 2024, Finloup secured €1 million for a dynamic entry into the electronic device leasing market. Amount raised in 2024: €117K Neuron AI is a tech startup, spun out of EU‐funded research, dedicated to transforming AI from lab to real-world impact. Founded in 2021, it unites software engineers, machine learning researchers, and market analysts to develop and commercialise cutting-edge AI solutions. Specialising in areas like predictive maintenance through automated ML (e.g., their OPTIPLANT project), IoT, AR, big data analytics, and optimisation, Neuron AI builds bespoke, enterprise-grade systems that boost operational efficiency. In 2024, Neuron AI was awarded a grant of up to €117,000 by the European Commission’s NGI TrustChain initiative.

EWOR, the operator-run fellowship that commits €60 million to early-stage founders, named Simon “Bobby” Dussart as Chief Operating Officer (COO).  Dussart was appointed CEO of Adjust in 2022 and guided mobile-analytics unicorn Adjust from Seed stage to more than €150 million in annual recurring revenue and a €1 billion exit, giving him the operator insight EWOR now embeds directly into its program. EWOR is a radically selective fellowship backing the world's top founders, built and led by unicorn builders.  It sifts through 35,000 + applications annually yet admits only 35 founders (0.1 per cent). Each selected founder receives a €500,000 cheque – one of Europe’s largest pre-seed tickets – and weekly one-to-one sessions with unicorn operators from SumUp, ProGlove and Adjust. Participants include Aeon and Zero One Creative.  Over the past two years, alumni have gone on to raise €1-11 million follow-on rounds within six months of graduation, underlining the effectiveness of EWOR’s operator-led model.  “Slide decks don’t build companies. Relentless operators do, and that’s exactly who our 0.1 per cent founders sit beside every week,” said EWOR Co-Founder Daniel Dippold.  “Simon lived every scaling trap – from the first thousand euros to nine-figure revenue. His repeatable playbook lets us serve fellows with the rigour of a co-founder, not the detachment of an investor.” That said, Adjust is not without controversy. In late November, Adjust laid off 304 employees. A company source told Tech.eu: "We got an automated, pre-recorded message from the CEO, and then we were all locked out of their devices. Such a nice, caring 'family'. Fortunately, everyone was paid appropriate severance.  As COO, Dussart will architect EWOR’s next growth phase.  “I know how to balance growth with profits with headcount with innovation, all while avoiding jail time for failing audits,” Dussart said.  “At EWOR we build a data-driven growth engine so founders can out-execute yesterday’s playbooks and build Europe’s next category leaders. EWOR exists to keep them two steps ahead of the market.” 

The Estonian provider of scalable smart building automation solutions, Bisly, has raised a €4.3 million funding round. The funding will support Bisly’s expansion in the DACH region and the UK, as well as the opening of a new office in Poland, one of the fastest-growing markets for smart building technology. In Estonia, Bisly already leads the market, with its automation systems installed in the majority of new residential apartments sold in 2024.  Buildings account for about 40 per cent of the EU’s total energy use, yet around 75 per cent are energy inefficient. As economic pressures rise and regulations tighten, demand for building automation is growing.  However, scalable, cost-effective solutions remain limited, as legacy systems are often expensive and unreliable. While transport electrification dominates climate discussions, improving building energy efficiency offers quicker and more affordable emission reductions. According to Bisly, each tonne of CO₂ avoided using its technology can be up to six times more cost-effective than achieving the same reduction through transport electrification. Bisly addresses this challenge with an AI-powered automation system that makes smart building technology both affordable and accessible. Its cloud-based platform simplifies sales, installation, and setup, cutting costs by up to 50 per cent compared to conventional systems. The company’s integrated hardware and software solution manages energy use and indoor environments in both new and existing buildings. At the core is Bisly’s patented digital twin technology, which creates a virtual model of each building and uses real-time data to continuously optimise performance. The system is also compatible with third-party devices, giving developers flexibility and avoiding vendor lock-in. Ants Vill, CEO and co-founder of Bisly, explained: The climate crisis hasn’t paused – 2024 was yet again, the hottest year on record. At a time of geopolitical and economic turbulence, we need practical solutions that work and have been deployed successfully in the field. Our technology helps solve the biggest source of energy consumption: buildings. Saving energy here doesn’t just help the environment, it frees up grid capacity for the next wave of energy demand, from AI data centres to mass-scale EV deployment.  Bisly’s system is currently in use across residential and commercial properties. Key features include centralised configuration, digital-twin building diagnostics, AI-powered BMS and reporting tools for developers and property managers. These capabilities have made Bisly’s platform attractive to a range of partners across the region, including Endover, Liven, Crown Estate, Bonava, Pro Kapital, Reterra, Hansab, Onninen and One-Gruppe, among many others.  The company works through a growing network of partners in each country and continues to expand technical and commercial support on the ground.  We want to thank our customers such as Endover, Nobe, Liven, Bonava, Nordecon, Onninen, and Oma Ehitaja, among many others, whose feedback and push for high reliability and affordability helped shape a product that’s now enabling mass adoption of smart building automation across Europe,  added Vill.  The round, led by 2C Ventures, included participation from existing investors, such as  Aconterra, Pinorena, the founders of Foxway, and SmartCap – Estonia’s state-backed VC fund, whose Green Fund investment is supported by the EU’s NextGenerationEU recovery plan.  Martin Koppel, Partner at 2C Ventures, commented:  We’re proud to support Bisly in scaling their technology across Europe and beyond. Buildings account for nearly 40% of global CO₂ emissions, and yet much of the sector remains underserved by digital innovation. Scalable, smart energy management is critical to reducing this footprint and Bisly’s platform brings that much-needed intelligence and efficiency to the built environment. The new capital will support go-to-market efforts in Germany, the UK, and Poland. Bisly already operates offices in Tallinn, London, and Berlin, and is expanding its engineering team to launch a wireless product line for retrofitting existing buildings.  Bisly is expanding its team with technical and sales hires in key markets and has named a Polish-speaking Head of Demand Generation ahead of opening a full office in Warsaw. The company tripled its core revenue last year and is on course for similar growth in 2025. Lead image: Bisly team | Photo: Uncredited

Britain's Ministry of Defence (MoD) has selected Australian firm Castlepoint Systems to provide services for automating data classification and reducing the risk of human error. The company is now headquartered in London. The selection marks Castlepoint’s first UK contract and comes on the heels of growing concern over the risks of mishandled sensitive information, most notably the 2022 Afghan data breach, which exposed nearly 19,000 personal records and drew sharp public criticism. “Securing this contract with the Ministry of Defence as our first UK account is a key milestone for Castlepoint, underscoring the critical importance of sophisticated data control for any organisation, not just national security,” said Rachael Greaves, CEO of Castlepoint Systems. “The MoD faces a complex challenge in managing vast and sensitive datasets in the knowledge that even a single case of data leak or loss can be catastrophic. I’m pleased that after undertaking a very thorough global search, Castlepoint was selected by MoD as the best solution to solve this problem." Castlepoint’s tech alleviates the reliance on manual data classification by personnel operating across high-stakes environments. Human error, particularly in labelling and handling unstructured data, has been cited as a weak link in national security. In Castlepoint’s system, data such as emails, reports, and documents are automatically assessed and given the correct security marking at the point of creation, based on both content and context. Importantly, the system is explainable, providing traceable justifications for each classification decision, an essential feature in meeting the UK’s ethical and legal standards for AI deployment. This contract signals Castlepoint’s formal entry into the UK and European markets, where demand for AI solutions in public sector data governance is intensifying. The company is already established in Australia and New Zealand, where its technology is used by two-thirds of the Australian Federal Government. “Castlepoint, with Explainable AI and true autoclassification at its core, can increase labelling accuracy and coverage without disrupting the essential work of MoD personnel,” Greaves said. “We are a trusted technology provider for public-sector organisations and enterprises in Australia and New Zealand, and having now established our global headquarters in London, we look forward to delivering our proven solutions to many more organisations in the UK.”

ChatGPT maker OpenAI is launching two open-source models (gpt-oss-120B and gpt-oss-20B), which it claims outperform similarly sized models on reasoning through complex tasks.The US frontier AI firm said the two models, called gpt-oss models, were its “most capable” open-source models to date. The launch of OpenAI’s open source models has been highly anticipated, along with the expected launch of GPT-5.OpenAI founder Sam Altman has previously tweeted about a delay to the release of its open-source model. Altman said: "We're excited to make this model, the result of billions of dollars of research, available to the world to get AI into the hands of the most people possible. As part of this, we are quite hopeful that this release will enable new kinds of research and the creation of new kinds of products.  "Going back to when we started in 2015, OpenAI's mission is to ensure AGI that benefits all of humanity. To that end, we are excited for the world to be building on an open AI stack created in the United States, based on democratic values, available for free to all and for wide benefit."The new open-source models mark the first open-source models for OpenAI since 2019 and put it up against the likes of Meta’s Llama models and Mistral’s AI models.  It also comes as Japanese open-source models increasingly gain traction. Unlike closed models, open-sourced models are available to developers, so they can build on top of them to improve their own products and tools. OpenAI said the two open-source models were trained using a mix of reinforcement learning and techniques informed by OpenAI’s most advanced internal models, including o3 and other frontier systems.OpenAI said the two open-source models outperform similarly sized open models on "reasoning tasks, demonstrate strong tool use capabilities, and are optimised for efficient deployment on consumer hardware". On the virtues of releasing open-source models, OpenAI said they give developers a “wider range of tools to accelerate leading-edge research, foster innovation and enable safer, more transparent AI development across a wide range of use cases”.The models also lower barriers for use in emerging markets, OpenAI said.The models will be available for developers in Europe and other countries to freely download and run locally.