• TechChill returns to Riga on March 25–27th 2026, gathering startups, investors, and tech leaders
• Over 2,000 participants are expected, including 300 startups and 250 investors from across Europe
• Founders Battle highlights early-stage startups pitching to investors for funding and visibility
• Side events and expert talks expand networking and ecosystem collaboration.

The Baltic region’s flagship startup and technology conference, TechChill 2026, returns to Riga, Latvia, on March 25–27th, bringing together founders, investors, and ecosystem leaders for three days of networking, startup pitches, and industry discussions.

Recognized as one of the leading startup events in the CEE, this year’s TechChill expects to welcome more than 2,000 participants, including over 300 startups, 250 investors, dozens of speakers, and media representatives from across Europe and beyond. The Conference resumes its role as a meeting point for innovators from the Baltic, Nordic, and Central and Eastern European regions, creating opportunities to exchange ideas, attract investment, and explore new partnerships.

The main conference program will take place on March 26–27th in Riga, featuring keynote talks, panel discussions, startup showcases, and networking sessions focused on scaling technology companies and navigating global markets.

Founders Battle to Spotlight Emerging Startups

One of the central highlights of the conference is the TechChill Founders Battle, the event’s signature startup competition. The contest brings early-stage startups to the main stage, where top 20 selected teams will pitch their products and business models in front of investors, mentors, and industry experts.

Participating startups receive exposure to hundreds of investors and ecosystem stakeholders attending the conference. The competition also provides founders with mentoring and feedback opportunities, helping them refine their pitches and strengthen their investment readiness before presenting on stage.

For many startups, the Founders Battle has become an important launchpad to secure funding, gain international visibility, and build strategic partnerships across the European startup ecosystem.

Side Events Expand Networking Across the City

TechChill officially kicks off on March 25th with a full day of community-driven side events taking place across Riga. These sessions are designed to encourage informal networking and knowledge exchange between founders, investors, and tech professionals.

Many of the side events are free and open to both TechChill attendees and the broader startup community, creating an accessible entry point for newcomers to connect with the ecosystem.

The side-event program includes a variety of formats, from expert workshops and industry discussions to community meetups and networking gatherings. Examples include sessions like StartSchool and Elevate, which offer practical masterclasses for early-stage founders, as well as thematic discussions like Baltic Startup Policy Forum, focused on startup-friendly, evidence-based policymaking.

Other events focus on specific sectors and communities, including Investor Day, From Qubits to Code: Scaling in Canada, and other sessions. In addition to professional sessions, informal networking formats, such as the Community Mixer, help bring together participants in a relaxed environment before the main conference begins.

Speakers and Topics

As one expects, TechChill brings together top-tier speakers and attendees from across the Baltics and beyond at its main event at Hanzas Perons. From aspiring tech students and future founders to seasoned entrepreneurs and investors, everyone will benefit from this exceptional chance to network and explore the latest tech trends showcased through real-world examples.

Having spoken to countless founders about why they came to TechChill in the first place and how they wanted to benefit from the event, TechChill received a unanimous answer – they needed fewer inspirational talks, and more guidance based on real-life experience.

‘We’ve always been here to support Baltic founders first and foremost, and the TechChill 2025 program reflects that,’ TechChill’s CEO Annija Mežgaile states.

Namely, this TechChill’s agenda will cover the following topics:

• AI and VC in the Baltics
• Dual-Use as the New Default
• EU-specific Startup and Scaleup Strategies
• Impact Product Journey
• State Institutions and Army’s Expectations from Innovation
• Trust in AI-Powered Commerce
• The complete agenda is available here.

The confirmed speakers include:

• ​Andra Bagdonaite, General Partner at FIRSTPICK
• Anna Baranenko, COO of Saint Javelin
• Andris Berzins, Partner at Change Ventures
• Viktor Bodnar, Chairperson of the Management Board at BlockBen
• Iveta Bruvele​, Founder and CEO at Wandoo Finance
• Julija JJ Jegorova, Founder and CEO at Black Unicorn PR
• Madara Knutina, Board Member at Startin.lv, Head of Operations at BADideas.fund
• Tina Luse, Managing Director at Fintech Latvia Association
• Ramona Miglane, Partner at Sorainen
• Kristaps Prusis, Founding Partner at Outlast Fund
• Kriss Pujats, Co-Founder and CTO at Gravity Team
• Tim Vaino, Investor at Practica Capital
• Agnese Veckalne-Lescinska, Board Member at Latvian Startup Association, Director of Operations at Edurio
• Linda Võeras, Investment Manager at Karma Ventures
• Ekaterina Zaharieva, European Commissioner for Startups, Research and Innovation
• and more. Check out the full list of speakers here.

In a Nutshell

• Where: Hanzas Perons and other venues across Riga, Latvia
• When: March 25–27th, 2026
• Tickets cost EUR 69–559 and are available here, use promo code ITKEY15 for 15% off
• More information: at the event’s website

The post TechChill 2026 Brings Baltic & CEE Startup Ecosystem Together This March appeared first on ITKeyMedia.

• AI infrastructure platform Backboard.io boasts top LongMemEval and LoCoMo benchmark scores, surpassing frontier models
• The startup was founded in 2025 to solve AI statelessness with persistent, portable memory
• Its strong CEE traction reveals pragmatic, cost-conscious developer practices
• Backboard.io’s future focus is the marketplace and ecosystem of AI tools to amplify open source innovation at scale

This February, Backboard.io—the Canadian startup developing an infrastructure platform of persistent, shared memory for AI systems across sessions and models—announced that it scored 93.4% overall accuracy of the LongMemEval benchmark and 90.1% accuracy on the LoCoMo benchmark. This marks the highest publicly reported result under consistent methodology so far.

Founders with Proven Track Records and the Structural Trigger

Backboard was founded in 2025 by Rob Imbeault (CEO), who previously co-founded and scaled Assent into a global enterprise SaaS company, and Jonathan Murray, a multi-sector founder who has built vertically integrated technology systems across hardware, software, and applied AI.

Their trigger to start Backboard.io was structural. As large language models became capable enough to power real workflows, it became apparent they were fundamentally stateless (i.e. lacking built-in memory of past interactions). Enterprises needed smarter models with durable, portable, auditable memory that could persist across agents, tools, and time. Having both built infrastructure businesses before, the founders recognized that the real bottleneck was not model intelligence but the absence of a shared memory layer. As multi-model routing and multi-agent systems accelerated, it became clear this was the moment to build foundational infrastructure rather than another surface-level wrapper. With a shared understanding of how to build systems that enterprises can actually rely on at scale, the duo believed they were the right team to deliver this solution.

Developer-Centric AI Infrastructure

They built Backboard.io as a developer-centric AI infrastructure platform that solves a core limitation of large language models — their statelessness — by providing a persistent, portable memory layer and unified API on top of 2,000+ models, letting applications retain context, preferences, and long-term knowledge across sessions and models without stitching together separate tools. It combines stateful memory, context management, retrieval-augmented generation, and multi-model orchestration into one stack so teams can build smarter, scalable AI systems faster.

Raif Barbaros, Partner at Mistral Venture Partners

‘Most people have seen that ChatGPT can “remember” things about you. But that memory is locked inside ChatGPT. Developers cannot get that same persistent memory through the standard API or move that state across different models or tools, but Backboard.io gives developers that superpower. We provide portable, persistent memory at the message level that works across models, not just inside one product. That means any AI tool can remember a user’s history, preferences, decisions, and context over time, even if the underlying model changes,’ Mr Murray explains.

Investor Confidence

This January, Backboard.io raised a sizable Pre-Seed round from Mistral Venture Partners, N49P, Garage Capital, and Developer Capital.

‘Backboard isn’t another AI wrapper. It’s foundational infrastructure. Rob has the vision and operational discipline to make Backboard the standard for how AI systems remember and retrieve information,’ Mistral Venture Partners’ partner Raif Barbaros states.

‘Rob is an exceptional founder who could be building anything but has smartly chosen the Memory Layer of the AI stack as his next challenge. When we looked into the competitors, and the technical decisions they made, it became clear to us the Backboard is more performant,’ Developer Capital’s CEO Jordan Steiner agrees.

Benchmarks and Practical Validation

Jordan Steiner, CEO Developer Capital

The capital injection allowed for the rapid expansion of the company’s capabilities, integration of thousands of models, and top scores on major AI memory benchmarks like LoCoMo and LongMemEval.

‘The dataset is designed to examine not just an LLM but any LLM-based system’s capabilities and blindspots in a fine-grained manner. Raw human performance is somewhere around 88 percent. Breaking the 90-percent threshold requires superhuman consistency in recall and reasoning. Most high-performing frontier models currently score around 80 percent on LoCoMo. The system built by Backboard.io is a far better attempt at simulating memory as it manifests in humans. It is practical, cheaper, scalable and doesn’t rely solely on brute-force LLM processing for answers,’ creator of the LoCoMo benchmark and research scientist at Databricks Adyasha Maharana comments in the Ottawa Business Journal article.

Architecture Philosophy

The investor confidence and the benchmark scores reinforce Backboard.io’s vision of memory as foundational infrastructure rather than a bolt-on feature. Emphasizing memory as a foundation and not a sideloaded plugin meant making early architectural choices that prioritized performance, stability, and cost efficiency over endless surface-level customization.

Adyasha Maharana, Creator of the LoCoMo Benchmark, Research Scientist at Databricks

‘You can always make something more configurable, but too many knobs often make systems slower, more complex, and more expensive. We chose a more opinionated core so memory is faster, more reliable, and affordable at scale. A simple analogy is a house. You can build plumbing and electrical systems into the foundation, or you can try to run extension cords and water lines along the walls after the house is finished. In theory, you can customize everything afterward, but it is messier, less efficient, and more prone to failure,’ Mr Imbeault explains.

Real-World Performance

It’s worth remembering that the mentioned benchmarks like LoCoMo or LongMemEval are academic constructs with the standard methodology, providing a clean validation. They still need to translate into real client ROI metrics — e.g., error rates in production workflows, developer productivity, latency in multi-session systems. With this in mind, Backboard.io optimizes memory at the message level, across real, evolving conversations with persistent state. The structural difference here is that the common approach of other products and open source approaches operate on prompts, not state, so they literally cannot do the same without controlling memory at the message layer.

‘Because our architecture manages memory natively at the message level, we can measure performance the way real systems behave in production, across sessions, across agents, and over time. That is not just a benchmarking advantage. It is a fundamental architectural difference. We built for real-world usage first. The fact that it also crushes traditional benchmarks is simply the outcome of designing for reality, not the test,’ Mr Imbeault emphasizes.

CEE Developer Engagement

The Backboard.io team mentions notable traction and strong developer engagement in Ukraine and Poland. According to Mr Murray, developers in these countries reveal a much stronger bias toward pragmatism. These developer communities were some of the first to actively experiment with multi-agent orchestration to drive token costs down. Instead of brute forcing tasks with a single expensive frontier model, they were coordinating multiple smaller assistants behind an orchestration layer to complete the same workflow faster and at a fraction of the cost, often using open source models.

‘That pattern showed us something important. In parts of Western developer ecosystems, there is still a tendency to anchor to a single model provider, likely because large platforms have distributed generous credits in exchange for early loyalty. In Central and Eastern Europe, we saw less attachment and more cost discipline. The focus was not brand alignment, it was performance per dollar and architectural efficiency. The distinct trend emerging is clear. The future is not one giant model doing everything. It is orchestrated systems of models, optimized dynamically for cost, speed, and task fit. That mindset aligns closely with our belief that portable state and model-agnostic memory are foundational, because orchestration only works well when memory is not locked into one provider,’ Mr Murray tells ITKeyMedia.

Open Marketplace and Tooling

He believes that CEE has a strong open source and academic culture, so Backboard.io’s focus should be on building infrastructure that amplifies that ingenuity. Over the coming months, the team is launching an open marketplace of tools built on Backboard, where developers and research teams can publish assistants, plugins, and workflows that others can fork, extend, and deploy. That creates a direct path from research prototype to reusable, production-ready infrastructure.

A model comparison tool is also scheduled for release, that allows developers to test multiple models against each other in real time, including open source models versus expensive reasoning models. The goal is to give builders empirical clarity on performance, cost, and memory behavior, so they can make informed architectural decisions without guesswork.

More broadly, we’re moving quickly to support open source projects like OpenClaw by releasing plugins and assistants as soon as possible, ensuring that Backboard integrates into the tools developers are already using. The intent is to bridge local research and open source momentum with persistent memory infrastructure that makes those innovations deployable at scale.

Isolation, Privacy, and Compliance

Backboard.io’s infrastructure involves shared memory across agents, which naturally raises questions about isolation and privacy. In multi-tenant, regulated environments (e.g. healthcare or finance), trust boundaries and compliance need to work with a unified memory layer. To address this, Backboard.io has memory scoped to an assistant_id, not to a global pool, so assistants are naturally siloed by default and you control what gets reused by choosing which assistant you attach a workflow to. Memories also persist across threads for the same assistant, which is powerful for continuity, but it means the clean trust boundary in regulated, multi-tenant environments is “one assistant per tenant, patient, account, or case,” so each tenant has an isolated memory store by design.

To share memory across agents, a developer opts into it by intentionally reusing the same assistant_id across agents or by building an explicit sharing layer on top with their own access controls and policy. On the compliance side, memory can be kept off when required, or managed directly by viewing, disabling, and deleting individual memories.

Continuity Across Models

Rob Imbeault, Founder and CEO at Backboard.io

Importantly, Backboard preserves continuity across model swaps because the memory layer is externalized and stateful. The conversation state, structured memory, and message-level relevance are not tied to any single model, which means developers can move from one provider to another without losing continuity.

‘Where continuity can get challenged is around context window management. When you move from a very large context window model to a smaller one, you have to be deliberate about what gets surfaced and compressed, otherwise you risk truncation or signal loss. We have an early solution in place that is working effectively today, and we’re close to pushing a more robust production tool that intelligently manages context packaging and relevance across model sizes. We’re also seeing multi-agent swarms built on Backboard naturally help with this, since specialized agents can operate within smaller windows while relying on the shared persistent state,’ Mr Imbeault comments.

Hackathons as Live Testbeds

Backboard.io organizes hackathons in collaboration with groups like AI Collective and regional partners as a live testbed for format, developer engagement, and follow-on conversion. They serve two purposes:

• On the professional side, they help organizations and senior engineers understand how to implement persistent memory in real workflows. In that setting, adoption is the metric. The organizers look at API key creation, assistant deployments, message volume, and whether teams move from a prototype into an active build within 30 to 60 days.
• On the university side, hackathons are a top of funnel GTM motion with awareness as its first objective. Blackboard.io can measure developer signups, credit activation, continued usage after the event, open source releases built on Backboard, and whether projects evolve into startups or sustained tools.

‘Participation is easy to count. Real impact is measured in retained builders and deployed systems,’ Mr Murray summarizes.

‘The model landscape is shifting constantly, open source is accelerating, and new agent frameworks appear almost weekly. What we can say with confidence is that Backboard will continue to focus on enabling developers first. We’re building for builders, listening closely to feedback, and shipping quickly in response to how people are actually using the platform,’ Mr Imbeault adds.

Future Ecosystem Expansion

Over the next few years, this likely means expanding the ecosystem layer around Backboard.io as much as the core infrastructure itself. The planned marketplace is a big part of that, creating distribution channels for tools built on Backboard.io so developers can turn prototypes into adopted products.

‘In regions like CEE, where there is strong open source and academic momentum, we see an opportunity to support local ingenuity with persistent memory infrastructure that makes those ideas deployable at scale. Our role is not to predict every trend, but to provide the stateful foundation that allows developers to adapt as the landscape evolves,’ the CEO remarks.

Backboard.io is redefining how AI systems retain and apply knowledge, turning inherently stateless models into stateful, agent-ready platforms capable of sustaining context, preferences, and workflows over time. Its traction in Central and Eastern Europe highlights a region that combines technical ingenuity with pragmatic, cost-conscious development practices, making it a natural proving ground for scalable AI infrastructure. By bridging local open source innovation with a persistent memory layer, Backboard.io is not just advancing AI performance—it’s empowering CEE developers to build the next generation of intelligent, adaptable systems.

The post CEE Developers Lead the Way with Backboard.io’s Portable AI Memory Infrastructure appeared first on ITKeyMedia.

At this year’s World Economic Forum, European Commission President Ursula von der Leyen publicly unveiled what has since been colloquially branded “EU Inc” (formally the 28th Regime), as part of a desire to strengthen the EU’s competitiveness and reduce fragmentation in corporate law. Created and vouched for by founders, investors and other private individuals and entities across the EU, EU Inc aims to create a single, optional EU-wide legal entity for some companies, that would function across all 27 member states under harmonised rules and a central digital incorporation system. Founders would, in theory, be able to register a company online in 48 hours and operate seamlessly across the EU when fully implemented.

Momentum and Past Initiatives

Support within the Commission and parts of the European Parliament is strong and visible. The 28th regime is included in the Commission’s 2026 work programme, with a legislative proposal scheduled for March 2026.

But the EU’s track record here is mixed. Attempts to harmonise corporate law have failed, such as the European Company (SE) (already in its third decade of existence), European Private Company (a 2008 initiative that was withdrawn due to little political support), or Single-Member Company (an initiative that still exists technically on EU institutions’ agenda). This occurs although EU Institutions themselves have highlighted that there is legal basis in the EU’s founding treaties for a wide alignment on multiple topics even outside corporate law, such as labour law, taxation, or insolvency. 

When Can We Expect EU-Inc to Become a Reality?

After the formal proposal in March 2026, the ordinary legislative procedure could take at least a couple of years before the regime is adopted. Following adoption, further technical work and coordination with member states would be required before EU-Inc becomes a practical tool.

As part of the legislative process, as parts of the initial proposal might be rooted in domestic law, pushback from Member States is likely where EU-Inc overlaps perceived sovereign domains.

EU-Inc’s success also hinges on EU institutions’ ability to build and manage a centralised digital infrastructure, the likes of which would be required to function across 27 jurisdictions.

The Reality To Which Founders and Investors Should Prepare

EU-Inc is framed as an optional regime, not a replacement of national legal forms. This optionality means that enforceability and uptake will depend heavily on how attractive and operational the new regime is in practice. Expected reduced administrative friction in incorporation and cross-border growth once EU-Inc is live shall be, most likely, balanced by issues such as tax treatment, labour compliance, and creditor rights that might still require national navigation.

The post EU-Inc and Why It Matters by Tudor Stanciu appeared first on ITKeyMedia.

• ZAKA VC backs Cirrus Therapeutics’ gene therapies for chronic blinding eye diseases, focusing on dry AMD
• The startup’s IRAK-M therapy targets upstream aging biology, restoring immune balance in the retina
• The new funding accelerates Cirrus Therapeutics’ development towards the clinic and expands next-generation ocular therapy pipeline

This February, the well-known Czech investor ZAKA VC (invested in Nephrogen, among others) announced that it’s joining the Seed round of Cirrus Therapeutics—the US-based biotech startup working on next-gen gene therapies for chronic blinding eye diseases. Other investors in the USD 11M round include ClavystBio, Polaris Partners, and SEEDS Capital.

Founders Bridging Science and Patient Care

Cirrus Therapeutics was founded in 2022 by two leaders whose combined expertise bridges cutting-edge science and frontline ophthalmic care. 

• Dr Ying Kai Chan (CEO) is an entrepreneur and translational scientist who previously served as Chief Scientific Officer at venture-backed Ally Therapeutics. His training spans neuroscience, immunology, and translational science, with deep expertise in genetic medicine platforms and therapeutic development. 
• Professor Andrew Dick (Chief Scientific Advisor) is a renowned physician-scientist with a background in biochemistry, immunology, and ophthalmology. He brings decades of clinical experience in inflammatory eye disease and diseases of the retina, including both medical and surgical retina. These perspectives ensure that the company’s scientific innovation is closely aligned with addressing real-world patient needs. 

Dr Ying Kai Chan, Co-Founder and CEO at Cirrus Therapeutics

Through a series of personal and professional experiences, Dr Chan came to appreciate just how devastating vision loss can be.

‘Vision is our primary interface with the world; when it deteriorates, independence, connection, and quality of life often decline with it. Yet many current therapies for vision loss are reactive—designed to manage late-stage complications rather than address upstream drivers of disease. We believed there was an opportunity to intervene earlier by restoring immune balance, with the goal of slowing or preventing progression of age-related macular degeneration and preserving central vision,’ Dr Chan tells ITKeyMedia.

The duo’s backgrounds uniquely positioned them to pursue that ambition. As Dr Chan transitioned from his previous Chief Scientific Officer role to Chief Executive Officer at Cirrus, his perspective broadened from platform innovation to company-building and long-term strategy.

In his turn, Professor Andrew Dick’s deep commitment to patients and the translational strength of his laboratory’s work in ocular immunology made clear that the science had real therapeutic potential.

Their shared motivation—to translate rigorous immunology into meaningful clinical impact—combined with emerging scientific insight, ultimately led them to co-found Cirrus. The goal is both straightforward and bold: to preserve sight, and restore it where possible.

Pioneering Multi-Pathway Gene Therapy

‘Cirrus’ bold approach focuses on pioneering novel yet well-validated biology from human genetics and aging data to deliver transformational impact in dry AMD, a disease of high unmet need. At ClavystBio, we are committed to backing Cirrus and its talented founders Kai and Andrew to advance this important therapy toward the clinic, and build its earlier stage pipeline,’ ClavystBio’s CEO Dr Khoo Shih shares.

Dr Khoo Shih, CEO at ClavystBio

Briefly put, Cirrus Therapeutics develops gene therapies for chronic blinding eye diseases, with an initial focus on dry age-related macular degeneration (AMD). What makes their approach stand out is that instead of targeting a single downstream pathway, they aim to address a root driver of retinal degeneration by restoring IRAK-M, a central regulator of aging-related inflammation in the eye.

Rather than targeting a single downstream disease pathway, this approach focuses on rejuvenating IRAK-M to influence multiple mechanisms involved in retinal degeneration. By addressing several interconnected biological pathways simultaneously, the therapy is designed to move beyond structural improvements alone. This multi-pathway strategy creates the potential for meaningful functional benefits, including improvements related to retinal metabolic function.

Unmet Medical Needs

Dry AMD, particularly its advanced form known as Geographic Atrophy, represents a significant and growing unmet medical need due to the eventual loss of central vision, and the team has identified that IRAK-M levels decline in the retina with age and decrease further in patients with dry AMD. Given that currently approved FDA treatments offer only modest effects without proven functional benefits, the company intends to advance its IRAK-M gene therapy program to address this gap.

Cirrus Therapeutics plans to deliver a one-time ocular gene therapy that could preserve or potentially restore vision by improving the underlying healthspan of retinal tissue. In other words, this is an ambitious bet on the idea of treating the biology of aging itself—rather than just one molecular domino in the cascade—and thus changing the trajectory of the disease rather than merely slowing it.

‘Cirrus clouds are high-altitude clouds and often indicate an approaching warm front. The name Cirrus Therapeutics signifies our aspiration and commitment to developing transformative therapies for chronic blinding diseases – a high and ambitious goal – while representing hope for millions of patients worldwide,’ Dr Chan explains.

‘The future of AMD treatment lies in understanding the immune system’s role in retinal health. At Cirrus, we are taking a bold, multi-pathway approach, leveraging bright minds in science to push the boundaries of what’s possible for patients,’ Professor Dick adds.

Academic Partnership and Clinical Rigor

Professor Andrew Dick, Co-Founder and Chief Scientific Advisor at Cirrus Therapeutics

Professor Andrew Dick, MD, is a clinician-scientist ophthalmologist internationally recognized for his work on retinal inflammatory diseases. He holds senior roles at both the University of Bristol and University College London (UCL), where he serves as Director and Duke Elder Professor of Ophthalmology at UCL Institute of Ophthalmology. Professor Dick co-directs the NIHR (National Institute for Health and Care Research) Biomedical Research Centre in Ophthalmology at UCL Institute of Ophthalmology. As such, these institutions remain central partners to Cirrus Therapeutics’ work to translate advanced ocular science into transformative therapies for dry AMD and geographic atrophy.

‘Cirrus was born out of deep academic collaboration, and staying close to our partners in Bristol and London is essential to how we innovate. Working side by side with Andrew and his lab keeps us grounded in rigorous science as we build a global company focused on protecting and restoring vision,’ Dr Chan comments.

Translational Studies and Pipeline Growth

As of today, Cirrus Therapeutics advanced its lead IRAK-M targeted AAV gene therapy program into translational studies, with the goal of moving into the clinic. The team also grew across scientific, strategic, and operational functions, strengthening the capabilities.

The startup is not yet disclosing its strategy for selecting its clinical trial populations across the spectrum of dry AMD stages. Ocular AAV delivery has been clinically validated in prior therapies (e.g. Luxturna), supporting the safety profile of this modality. The localized and low-dose approach is intended to minimize immune-related risks while maintaining therapeutic efficacy.

Beyond its lead program, Cirrus Therapeutics is developing first-in-class gene and cell therapies to preserve sight and extend ocular healthspan in patients with chronic blinding diseases. The company plans to apply its scientific expertise to multiple stages of dry AMD, including earlier intervention to prevent degeneration and next-generation treatments for advanced disease characterized by both degeneration and dysfunction. This strategy may also extend to other chronic blinding conditions with significant unmet medical need.

Shaping the Future of Vision Restoration

Ján Kasper, Co-Founder and Managing Partner at ZAKA VC

‘AMD represents a massive market with significant unmet need. Paired with Cirrus’ world-class science, strong execution, and backing from top-tier biotech investors, this made Cirrus a compelling fit for us. We are excited to support the team as they advance a truly differentiated approach to treating blinding diseases’, ZAKA VC co-founder and managing partner Ján Kasper concludes.

The newly raised funding is meant to accelerate the progression of Cirrus Therapeutics lead candidate toward IND-enabling development activities. More specifically, this will help the team complete the preclinical work required to support a regulatory submission. In parallel, the company will continue developing its expanding pipeline of next-generation therapies for ocular diseases.

Cirrus Therapeutics is betting on a bold shift in approaching vision loss—moving upstream to target the aging biology that drives degeneration rather than managing its late-stage consequences. If successful, its IRAK-M–based gene therapy could redefine what durable treatment means for dry AMD and potentially reshape the broader field of ocular medicine. ZAKA VC’s involvement reinforces international investor confidence in this vision and signals that transformative biotech innovation emerging at the intersection of immunology and gene therapy is attracting increasing global backing altogether.

The post ZAKA VC Joins Cirrus Therapeutics’s Seed Round, Transforming Vision Care with Next-Gen Gene Therapy appeared first on ITKeyMedia.

• Hospitality platform Mews brought in USD 300M Series D from EQT Growth, Atomico, and other investors
• The funding supports AI integration across hotel operations, reducing staff workload and improving experiences
• Enhancing automation and analytics capabilities, Mews aims to set new hospitality standards with AI-driven, human-centric solutions

This January, Mews—the well-known Czech-founded hospitality tech unicorn—raised its Series D round of USD 300M. The lead investor of the round is EQT Growth, joined by the famous VC firm Atomico (invested in Dexory, among others) and HarbourVest Partners, alongside Mews’ long-standing supporters Kinnevik, Battery Ventures, and Tiger Global.

Founding Vision and Rapid Expansion

Mews is a cloud-based hospitality management platform founded in 2012 and headquartered in Amsterdam. The company was established by Richard Valtr and Matthijs Welle with the goal of modernizing hotel operations through fully cloud-native software.

Mews launched its first property management system in 2013, targeting independent hotels that wanted an alternative to legacy on-premise hotel software. Early traction came from European boutique and lifestyle hotel groups, helped by the platform’s API-first architecture and integrations with hundreds of hospitality tools.

The company began scaling rapidly after major funding rounds throughout the 2020s. Those investments helped Mews expand internationally and accelerate product development.

Strategic Acquisitions to Strengthen the Ecosystem

Michael Coscetta, President at Mews

Mews also pursued growth through acquisitions, including companies such as Hotello and Cenium, strengthening its presence in North America and enterprise hotel segments. The company’s latest acquisitions include Flexkeeping, the hotel operations management platform, and DataChat, the leading generative AI analytics platform.

‘Our acquisitions strengthen our operating system and compound customer value. Culture fit is a critical part of such decisions. Cultural consistency comes from a clear mission, shared product principles, strong ownership, and integration focused on impact rather than just deal volume,’ Mews’ President Michael Coscetta comments.

Global Reach and Platform Capabilities

Laura Connell, Partner at Atomico

Today, Mews serves thousands of hospitality properties across more than 80 countries, including hotels, hostels, serviced apartments, and hybrid accommodations. Its platform covers property management, payments, guest experience, and operational automation, positioning the company among the leading modern alternatives to legacy hospitality systems.

‘Matt and Richard have built a category-defining platform with the depth, pace of innovation and global reach required by modern hospitality. The most ambitious builders in hospitality are focused on delivering ever-improving experiences for their end consumers, and they need technology that can keep pace with rising expectations around speed, service and personalization. Mews is the key enabler for the future of hospitality, and the team is well on their way to building a generational company,’ Atomico’s partner Laura Connell states.

EQT Growth’s partner Kirk Lepke shares that he had the pleasure of getting to know the Mews team several years ago and witnessed the company go from a bold vision to an organization delivering at scale. According to him, being one of the world’s largest industries, hospitality still has its core systems lagging decades behind, and Mews is creating a modern technology standard, an AI-enabled hospitality operating system that helps solve the fragmentation in the industry.

Accelerating AI-Driven Transformation of Hospitality

Kirk Lepke, Partner at EQT Growth

The new USD 300M investment is meant to expand Mews’ investments in AI, embedding agent-driven systems across the platform to automate complex workflows, reduce cognitive load for staff, significantly improve the guest experience and accelerate how products are built and deployed.

‘With EQT Growth joining in addition to new investors Atomico and HarbourVest, we have the backing to continue moving faster than anyone else in the industry. We are engineering an operating system that is changing how hoteliers interact with their guests. Mews exists to handle the operational complexity so hoteliers can focus on what matters: making hospitality even more fun, profitable, and fulfilling,’ Mr Welle states.

Matthijs Welle, CEO of Mews

More specifically, this funding accelerates Mews’ multi-year transformation into the AI-native operating system for hospitality. Over the next 12 to 18 months, the company intends to expand AI across revenue, operations, and guest workflows, with copilots embedded directly into the tasks hotel teams already perform every day, such as rate recommendations, task routing, guest communication support, and operational summaries. Over the following 18 to 36 months, Mews expects to move toward more agent-driven orchestration across departments, where the system can coordinate actions across workflows with clear human oversight.

Assessing AI Success and Operational Impact

The success of such transformation can be measured through a combination of operational, adoption and commercial KPIs, including adoption rates of AI-assisted workflows, manual hours saved, time saved to train new staff, self-service completion rates across the guest journey (such as check-in and check-out), workflow resolution times, operational efficiency gains, and revenue or margin uplift for customers where applicable.

To ensure that rapid AI integration improves hotel workflows without introducing new operational complexity for staff, Mews adheres to the core principle that AI must remove work and make everyone’s jobs easier instead of creating more tools. With this in mind, the platform is embedding intelligence directly into the workflows hotel teams already use, with clear human oversight, guardrails, and the ability to override automated actions where needed. Because Mews is a unified platform across PMS, payments, housekeeping and revenue management, AI can orchestrate within one system instead of adding fragmented layers that increase switching costs and training burden.

‘We also sequence implementation carefully. We prioritize high-frequency, low-risk workflows first, then expand automation as reliability and customer confidence are proven. If a capability adds steps or complexity for staff, it is not the right implementation, and we are working with any of our customers to design these use cases,’ Mr Coscetta assures.

Privacy, Compliance, and Human Oversight

Seeing how hospitality is among the industries where trust is foundational, Mews builds its AI approach around privacy-first design, GDPR compliance, human-in-the-loop controls, and clear boundaries for automation.

‘We focus on operational insights in aggregate, not personal profiling, and we continue to invest in platform security and governance as hotels rely on Mews as their system of record. We want hospitality to remain human, so our goal is for AI to enhance human productivity, not to remove humans from the experience,’ Mr Coscetta specifies.

Risks, Operational Challenges, and Client Success

He names key risks that the company anticipates over the next 2-3 years, both technically and operationally, in realizing its AI-native vision. They include ensuring AI remains safe, reliable, and explainable in real-time hotel operations, and avoiding ‘bolt-on’ complexity that creates more work instead of less.

‘Another risk is uneven data quality and operational readiness at the hotel, which can affect how quickly automation delivers value. That is why rollout sequencing, integration discipline and customer enablement are just as important as model capability. Operationally, the challenge is adoption and change management across diverse markets and property types, while maintaining the right balance between automation and hospitality’s human touch,’ Mr Coscetta adds.

He is convinced that, alongside financials, client success is human. It can be measured by what Mews can observe directly through platform and workflow outcomes, such as reductions in administrative workload, faster onboarding times, improved workflow completion times, and higher levels of guest self-service adoption. The team combines those indicators with customer feedback signals, where available, to understand whether teams are experiencing lower cognitive load, better day-to-day usability and more time for guest-facing work. Essentially, the goal remains to help teams spend less time on screens and more time on hospitality.

Market Validation and Accessible Automation

Richard Valtr, Founder at Mews

‘Hospitality is the business of experiences. The validation for our product from the market is clear, in both the US and Europe, and it is great to see how we are now powering ahead of any other hospitality company in terms of AI and agentic hospitality. It’s an exciting time to reinforce our vision of making Mews hotels the most profitable in the industry,’ Mr Valtr summarizes.

Importantly, Mews specifically focuses on making its automation capabilities accessible through embedded copilots, pre-set workflows, simpler onboarding, and an open ecosystem that reduces IT burden not only for large hotel groups. For smaller operators in particular, this means specific practical adoption paths, easier setup, preconfigured workflows, and solutions that don’t require large in-house technical teams. The vision is that independents should run with the same intelligence and efficiency as the largest groups, without legacy complexity.

The USD 300M Series D round positions Mews to accelerate its transformation into a truly AI-native hospitality operating system, embedding intelligence across operations, revenue, and guest workflows. By combining global scale, advanced automation, and human-centric design, Mews aims to level up hotel operations making them more efficient, profitable, and seamless for both large groups and independent operators. The funding reinforces the company’s ambition to set a new standard for the future of hospitality, where technology enhances human creativity and guest experience rather than replacing it.

The post AI-Native Hotels: Mews’ USD 300M Round Powers Hospitality Transformation appeared first on ITKeyMedia.

• A research by NordProtect identifies card cloning and cryptocurrency scams as most effective
• Social engineering, AI, and social media amplify scams despite technical safeguards
• A secure future requires public education, verification standards, and proactive cybersecurity responsibility

This January, the world-famous Lithuanian-born cyber security company NordProtect conducted its research on the most effective scams in the United States. Card cloning or skimming and cryptocurrency fraud led the top 10 ranking.

NordProtect and Its Methodology

A subdivision of Nord Security, NordProtect is a US-focused identity theft protection service designed to help people detect, prevent, and recover from digital fraud. It monitors credit activity, scans for data leaks and suspicious use of personal information, and alerts users to potential identity theft before the damage spreads. Beyond monitoring, it also provides recovery support and guidance.

NordProtect commissioned the survey on the most popular scams at Cint, which conducted it in December 2025. The survey’s target group comprised 1,004 online users in the USA aged 18-74. Quotas were placed on age, gender, and place of residence.

Specifics and Demographics

According to NordProtect’s research, the top 10 most effective scams (with percentage of victims as a share of encounters indicated) are:

1. Card cloning or skimming (45%)
2. Cryptocurrency scams (44%)
3. Romance scams (38%)
4. Calls from fake customer support or banks (37%)
5. Fake QR codes (37%)
6. Fake app scams (35%)
7. Impersonation of a friend or relative asking for money (35%)
8. Fake loan offers (33%)
9. Fake online shops (33%)
10. Fake product or service listings on legitimate platforms (32%)

Despite their high effectiveness, card cloning or skimming and crypto scams aren’t the most common. The most frequently encountered scams are phishing emails (31%), calls from fake customer support or banks (20%), and fake job offers (20%).

A comprehensive demographic analysis of scam victims was not conducted, but the survey did investigate the prevalence of specific scams across different age and sex groups. The findings indicate that internet users between 18 and 24 years old encounter scams involving fake online stores and fraudulent paid services more often than other age brackets. Regarding financial scams such as credit card cloning and cryptocurrency fraud, victims are most frequently 25-34 years old.

Classic VS Emerging Fraud Techniques

Tomas Sinicki, Managing Director at NordProtect

‘Card cloning and cryptocurrency scams aren’t just effective — they’re ideal for criminals. Crypto scams are popular because they’re high payoff and low risk — once funds are sent, they’re probably gone forever. Card cloning is the ‘classic rock’ of fraud — it’s old, but it still works thanks to the large number of payment cards in circulation and people not checking statements regularly,’ NordProtect’s managing director Tomas Sinicki states.

At that, the increasing adoption of contactless card payments is rendering traditional card cloning fraud less viable. Consequently, fraudsters are evolving their methods to steal and exploit payment card details through digital channels:

• One prevalent method involves stealing card data via phishing, malware, or social engineering attacks.
• A second, increasingly popular method uses mass-scale spam emails or text messages that direct users to fraudulent websites impersonating legitimate entities like banks or courier services where victims get prompted to enter their card details and one-time passwords.

In both scenarios, the captured data is then loaded onto a burner phone or into a digital wallet. This allows the fraudster to make purchases or withdrawals, achieving the same outcome as physical card skimming. This technique is often referred to as a ‘digital cloning equivalent’ and it transforms the way classic card cloning works.

‘Card cloning and crypto scams still require knowledge and effort, while phishing scams have the lowest barriers to entry for beginner scammers. Phishing is like the flu of the digital world — it’s everywhere, and it’s easy to catch,’ Mr Sinicki comments.

Cryptocurrency Fraud Dynamics

Specifically, the technical nature of cryptocurrencies also contributes to their popularity among scammers. The mostly irreversible nature of blockchain transactions means that once funds are transferred, they are exceptionally difficult to recover, making it a low-risk crime.

A question arises whether banks and crypto platforms should bear more liability when users fall victim or such liability risks encouraging reckless behavior. Mt Sinicki points out that in the majority of these incidents, financial institutions and platforms serve merely as the tools used by the victims themselves. Defining the liability of these institutions is challenging, because technically victims initiate the fund transfers.While financial institutions must take relevant steps to protect their customers, the public is strongly encouraged to exercise caution and vigilance, adopt preventative security measures, and maintain sufficient cyber insurance coverage as a safety net.

Best Practices for Protection

A layered approach — combining proactive monitoring, strong digital hygiene, and a well-defined response plan — provides the most effective protection. If personal data is suspected to be compromised, NordProtect recommends:

• Changing passwords. Email, banking, social media, and any accounts storing payment information should be prioritized. Old passwords are not to be reused. If one account is potentially compromised, similar passwords across other services should be considered at risk. Strong, unique passwords are essential.
• Enabling multi-factor authentication. Activated MFA significantly reduces the likelihood of unauthorized access, even if a password got compromised.
• Reviewing bank and card statements. All transactions should be carefully checked, and any unauthorized charges reported immediately.
• Monitoring credit. New accounts, unexpected credit inquiries, or sudden changes in credit scores may indicate identity theft and should be addressed promptly.
• Being cautious with unexpected communication. Emails, calls, or messages requesting financial information or urgent action should always be verified independently.
• Checking public records. Identity misuse can extend beyond financial fraud. Any indication that one’s name is linked to unfamiliar legal or criminal activity requires immediate attention.
• Considering identity theft protection services. While no technology can fully eliminate social engineering risks, there are reputable services that provide structured monitoring and recovery support for victims of online fraud.

Romance Threats and Tips for Protection

It’s important to note that romance scams are already among the fastest-growing categories of online fraud and rapidly developing voice- and image-cloning technologies has already impacted the effectiveness and scale of romance and impersonation scams and will definitely make scams more persuasive in the next few years.

According to Low Cost Detectives, romance scams have increased from USD 547M in 2021 to over USD 1.3B in the United States in 2024, with victims losing an average of USD 15K. Victims aged over 50 are the primary targets, nearly half are affected.

Mr Sinicki suggests some tips on avoiding romance and impersonation scams and protect oneself online:

• Remain skeptical but respectful. Stay cautious without undermining the relationship by doubting every word. If an online acquaintance takes offense, it is recommended to explain the need for caution. A genuine person will usually understand, while a scammer may respond with anger or defensiveness.
• Seek a second opinion from someone trusted. Before becoming emotionally invested, consult a friend or family member who may notice warning signs that were previously overlooked.
• Stay vigilant, even when initiating contact. Starting a conversation does not eliminate risk, as scammers can still manipulate the situation.
• Analyze videos carefully. AI-generated videos may reveal subtle flaws, such as unnatural facial movements or blinking, unusual expressions, blurred facial edges, inconsistent skin tones, or irregularities in teeth and fingers.

Social Media’s Role in Modern Fraud

Naturally, social media additionally enhances both the reach and impact of fraudulent activities significantly:

• First, oversharing on social media unintentionally provides fraudsters with information they may misuse. Social media serves as a vast repository of voice and image data, which allows criminals to create sophisticated AI-generated content for impersonating acquaintances and family members. The wealth of personal information and details from private events available on these platforms enables highly customized and convincing scams.
• Second, social media functions as a direct distribution channel for certain types of fraud, such as the promotion of fake e-commerce websites.

Biometry VS Systemic Identity Verification

‘Biometric authentication unquestionably enhances personal security and is effective at mitigating specific fraud types. Therefore, its implementation is highly recommended where available. Nevertheless, it is crucial to recognize that fraudsters adapt their tactics in response to new security measures. As noted previously, the most prevalent scams today do not exploit technical vulnerabilities but rather rely on social engineering to manipulate human psychology where technical solutions have limitations,’ Mr Sinicki adds.

Olga Voloshyna, Chairperson of the Committee on IT and Cyber Security of the German-Ukrainian Chamber of Industry and Commerce, CEO at Silvery LLC

Olga Voloshyna of the Committee on IT and Cyber Security of the German-Ukrainian Chamber of Industry and Commerce is convinced that the most vulnerable point in today’s digital environment is the system of identification and verification. It is precisely through this layer that fraudsters are able to scale their schemes so easily: most of them are built not on technical breaches, but on the skillful creation of someone else’s digital identity and the exploitation of trust.

‘The situation can be changed by shifting the focus from reacting to incidents to proactively confirming authenticity. If users could see how reliable a sender is before any interaction takes place, the opportunities for manipulation would decrease significantly. It’s about creating a shared mechanism for confirming legitimacy that would function across different digital environments—from messengers to online banking. If platforms begin to adopt common source-verification standards, the mass creation of fake accounts would become significantly more difficult and costly. Rethinking how authenticity is confirmed online may be the step that truly curbs the scale of modern fraud schemes,’ Ms Voloshyna shares.

Digital Fraud as a Structural Feature

What’s apparent is that changing today’s digital ecosystem to reduce scam effectiveness at scale is a multifaceted question with complex considerations. Many advanced tools heavily used by scammers, such as AI and social media, also provide significant benefits to the digital ecosystem and are therefore not feasible to eliminate. At that, criminals are inherently adaptive and will continuously seek new vulnerabilities within any changing digital environment. A complete redesign of the ecosystem in search of a perfect solution is unlikely to yield the expected results.

Therefore, Mr Sinicki believes that the most sustainable and instantly actionable solution lies in continuously educating society about emerging cybersecurity threats and scams, understanding their mechanics, and utilizing effective tools to protect oneself and provide support in the event of an incident.

The findings of the NordProtect research underscore that digital fraud is no longer a marginal threat but a structural feature of today’s interconnected economy. As scams grow more sophisticated—blending social engineering, AI, and financial technology—the line between technical vulnerability and human manipulation continues to blur. Addressing this challenge requires sustained public awareness, stronger verification standards, and shared responsibility across institutions, platforms, and individuals to reduce the scale and impact of modern fraud.

The post NordProtect Study Highlights Risks and Defenses in Modern Cyber Fraud appeared first on ITKeyMedia.

In my view, the most vulnerable point in today’s digital environment is the system of identification and verification. It is precisely through this layer that fraudsters are able to scale their schemes so easily: most of them are built not on technical breaches, but on the skillful creation of someone else’s “digital” identity and the exploitation of trust.

Creating an account and impersonating another person or company is relatively simple today. Fraudsters actively take advantage of this: they create fake profiles, send messages on behalf of supposedly well-known services, and construct entire chains of deception. This is not a problem of a single social media or messenger. It is a consequence of how most online interactions are structured.

When a person is unsure with whom they are communicating, any message can turn out to be dangerous. A link may be fraudulent, customer support may be fake, and a familiar contact may in fact be a bot. In such an environment, trust forms quickly, while verification is almost nonexistent.

Yet the situation can be changed by shifting the focus from reacting to incidents to proactively confirming authenticity. If users could see how reliable a sender is before any interaction takes place, the opportunities for manipulation would decrease significantly.

This is not about complicating life for users or introducing total control. Rather, it is about creating a shared mechanism for confirming legitimacy that would function across different digital environments—from messengers to online banking. Such a system doesn’t have to be centralized. A federated model is possible, where different providers verify individuals or services according to common standards. In higher-risk scenarios, such as financial transactions, this role could be performed by government eID systems or commercial identity services. The key principles should be voluntariness, data minimization, and respect for privacy.

If platforms begin to adopt common source-verification standards, the mass creation of fake accounts would become significantly more difficult and costly. Fraud would not disappear, but it would lose its simplicity and scale. Human vigilance would still matter, but technical tools could fundamentally reshape the logic of digital interaction: instead of relying on assumptions about whom to trust, users would receive a clear and verified signal. Rethinking how authenticity is confirmed online may be the step that truly curbs the scale of modern fraud schemes.

The post The Weakest Link Online: Why Digital Identity Is the Real Security Gap by Olga Voloshyna appeared first on ITKeyMedia.

• Demoboost brought in EUR 2.8M to scale its AI-powered B2B demo automation platform
• The startup’s founding team combines sales, tech, and buyer experience for a unique solution perspective
• The platform transforms interactive demos into behavioural data, shortening cycles and improving conversion
• Following the investment, Demoboost’s plans focus on Demo-Led Revenue and global expansion, including Series A in 12–18 months

This February, Demoboost—the Warsaw-based startup providing an AI-powered B2B software demo automation platform—raised its Seed round of EUR 2.8M. The co-lead investors of the round are Digital Ocean Ventures and Rafał Brzoska’s family office RIO, joined by the well-known Polish VC firm bValue (invested in XVision, among others) and several angel investors from the Vestbee Angels syndicate.

Founders’ Unique Demo Perspective

Demoboost started in 2021 with Paweł Jaszczurowski, Kamil Smuga, and Piotr Zesiuk, who were all deeply involved in selling Salesforce in Poland, one from the vendor side, one as a partner, and one as a client. They experienced firsthand how central demos were to closing deals, and at the same time how manual, repetitive, and inefficient the process was. That was the original trigger around which they began building their solution.

Anna Decroix joined shortly after as the CEO. Through her Polish network and startup mentoring work, the team reached out while they were building the product. What started as mentoring quickly became something much bigger. Before Demoboost, Ms Decroix was the CMO at Dyson in Australia where she led digital transformation and scaled e-commerce. She got to buy a lot of complex software and saw how painful and slow the buying journey could be from the buyer’s side.

Lech Kaniuk, Angel Investor, Member of Vestbee Angels, Chairperson at Sunroof

As such, what gives the co-founders confidence that they’re the right team is that they all understood the demo problem from different angles: sales leadership, tech and partnerships, and buyer experience. They realised demos were a major bottleneck in B2B software: too manual, too expensive, not scalable, and not accessible when buyers actually wanted them.

‘I know from experience the difference between a product that looks good in a presentation and one that actually delivers results. Demoboost falls into the latter category. I’ve already connected them with SunRoof and will be introducing them to other companies with which I’m involved,’ Vestbee Angels’ member and SunRoof’s chairperson Lech Kaniuk comments.

Demoboost’s Three Phases of Growth

Ms Decroix describes Demoboost’s journey so far in three phases:

• Phase one was bootstrapping. The team built an MVP and started selling with a very basic version of the product. It wasn’t perfect, but it proved something critical: there was real demand for a better way to deliver product demos. That validation gave the co-founders the confidence to keep building.
• Phase two started with Demoboost’s first funding round (EUR 1.7M, October 2022, led by b2Venture). At that point, the team became much more analytical about defining true product-market fit. Initially, they were selling broadly to anyone who saw value in demo automation. After analysing usage, retention, expansion and deal sizes, a clear pattern emerged: the strongest traction was with sales and presales teams in complex B2B SaaS environments. Around 2024, Demoboost doubled down on that ICP, refined our positioning around revenue teams.
• Phase three has been scaling with conviction. Once Demoboost’s product and go-to-market was aligned around that persona, growth accelerated significantly. Over the past years, the company tripled its ARR, increased average revenue per account by 50%, and were recognised in Gartner’s Demo Automation Market Guide as the most robust demo automation platform in the market. That growth was an important signal to both the co-founders and investors that the startup had identified what works and wasn’t experimenting anymore.

‘One of the biggest challenges along the way was the rapid rise of AI. Some investors questioned whether demo automation would become obsolete. Instead of debating the theory, we embedded AI deeply into our platform — particularly in demo creation and personalisation — and actually saw strong traction, especially from AI-native companies. Each phase built on the previous one: proving demand, defining focus, and then accelerating growth,’ the CEO recalls.

AI-Powered Demo Platform to Solve Concrete SaaS Sales Inefficiencies

To reiterate, Demoboost provides B2B software companies with a platform to create, personalize, and share interactive product demos at scale using AI. It allows revenue teams to track how buyers engage with those demos, turning clicks, views, and interactions into real-time intent data. In practice, it transforms demos from one-off sales presentations into a measurable revenue intelligence layer that helps teams prioritize deals, reduce manual work, and shorten sales cycles.

Piotr Widacki, Partner at Digital Ocean Ventures

‘Demoboost addresses a very concrete inefficiency in B2B SaaS sales. Too many teams still treat demos as presentations, not as a source of data. Demoboost turns the demo stage into a signal layer that helps sales teams understand real buying intent and focus effort where it matters. Combined with strong enterprise traction and a team that understands both sales and product deeply, this was a clear investment decision for us,’ Digital Ocean Ventures’ partner Piotr Widacki tells ITKeyMedia.

Global Adoption and Measurable Impact

The platform is used by global B2B SaaS leaders including AMCS, AvePoint, Bosch, Boost.ai, Celonis, and Hitachi. The startup reports that client companies have reduced unqualified sales calls by up to 74%, lowered customer acquisition costs by 28%, and removed demo wait times altogether. Interactive demos empower sales teams to lead introductory conversations without presales support, making early buyer interactions more informed and efficient, while reducing top-of-funnel workload and shortening sales cycles by a reported average of 32%.

Adam Bartkiewicz, Partner at Digital Ocean Ventures

‘Demoboost has a clear product-market fit, and they’ve been winning large, global customers from day one. The team has a deep understanding of their target audience, and it shows in product adoption and retention. With this round, they’ll accelerate international scaling and go after enterprise clients. What excites us most is that the team understands how B2B sales is evolving in the AI era – and Demoboost is evolving with it. We’re thrilled to support them as they scale globally and help shape what modern B2B sales looks like,’ Digital Ocean Ventures’ partner Adam Bartkiewicz states.

Behavioural Data Powering Decisions

It’s important to note that clients don’t ‘supply’ behavioural data to Demoboost. It is generated by demo viewers themselves as they interact with the experience: which paths they choose, on which screens they spend more time, which calls to action they click, and whether they share the demo with colleagues. These behavioural signals are collected within the demo environment and handled in line with each client’s privacy policies and setup. Demoboost  provides the ability to turn this interaction data into actionable insights. Teams can use it to:

• qualify leads more effectively (understand which use cases or features a prospect is genuinely interested in);
• map decision stakeholders (see who engages with the demo, how it spreads within the organisation, and identify additional influencers or decision makers involved in the buying process);
• prioritise sales follow-up (identify engaged stakeholders and focus effort on the most active opportunities);
• improve onboarding and customer success (see what new customers explored before purchase and tailor onboarding accordingly);
• and inform product and GTM strategy (aggregated insights can highlight which features attract the most attention).

In short, customers capture and interpret real buying-journey signals that were previously invisible, and Demoboost doesn’t retain and own this data.

Zbigniew Lukasiak, Angel Investor, Member of Vestbee Angels

‘First of all, Demoboost’s own demo was enthusiastic, open-minded and well-researched. And I believe that they are working with a very interesting problem. They essentially make a simplified copy of the software to be demoed — but there is the tension between simplification and showing off the software capabilities. If the demo is great, then maybe it should feed back into the software interface design? The extra point is about how AI is excellent for exploring these ideas and automating this work,’ Vestbee Angels’member Zbigniew Lukasiak remarks.

Responsible AI Personalisation

Furthermore, AI-driven demo personalisation in Demoboost is designed to be guided by user choice. Personalisation starts with explicit input, either the viewer or the salesperson selects who they are, what role they play, or which problems they’re trying to solve. Based on that, the platform surfaces the most relevant content paths.

‘We don’t infer sensitive attributes or make opaque assumptions about the buyer. Importantly, humans stay in control. Revenue teams design the demo structure, the available content options, and the personalisation logic themselves, and the AI helps execute that logic at scale,’ Ms Decroix assures

.Viewers, in their turn, actively choose their journey rather than being silently segmented behind the scenes. They can often explore beyond their initial path, which makes their experiences less narrow and restrictive. In short, Demoboost’s approach helps teams personalise responsibly, without reinforcing bias, focusing on relevance through explicit choice and human oversight rather than opaque AI profiling.

Revealing the Hidden Stakeholders and Deal Health Signals

Anna Decroix, Co-Founder at CEO at Demoboost

Ms Decroix shares one common example that comes from analysing how demos spread inside target accounts. In many sales processes, teams interact directly with only a handful of stakeholders — often the visible ‘tip of the iceberg.’ Critical decision influencers from IT, adjacent departments, or C-level leadership frequently don’t attend meetings, and their priorities remain unclear. By observing behavioural demo data, one client discovered that after initial calls, the demo was consistently viewed by additional stakeholders to whom they hadn’t spoken. This insight prompted the sales team to enrich their pitch: they added dedicated content paths for technical validation and executive value earlier in the process, and proactively equipped their champions with materials tailored to those hidden stakeholders. The result was a more complete, account-wide narrative and smoother decision alignment — simply because the demo data revealed who was really involved in the buying decision, not just who showed up to meetings.

Beyond sales efficiency, one of the most valuable and sometimes unexpected insights comes from the scale of engagement within an account. Clients often discover that the number of stakeholders interacting with a demo is a strong signal of deal health. When demos get viewed repeatedly and shared across multiple roles, the likelihood of closing the opportunity increases significantly. This shifts the strategic focus from simply running demos efficiently to identifying truly winnable deals.

‘Sales teams can prioritise opportunities where engagement is spreading across the organisation, while proactively supporting accounts where interest hasn’t yet broadened — for example by equipping champions with tailored demo paths or content for additional stakeholders. In that sense, demo engagement data becomes not just a sales metric, but a strategic signal helping teams focus their time and resources where they have the highest chance of winning,’ Ms Decroix explains.

Global Insights into Buyer Behaviour

As Demoboosts expands internationally, one of the key lessons that the team learned is that the core buying behaviour is surprisingly consistent, but how engagement happens can vary slightly by region and seniority. Across markets — Europe, the US, Australia, and beyond — the structural pattern remains the same: buying groups are large, and most of the evaluation happens independently of sales conversations. Buyers actually spend only a small portion of their journey in direct contact with vendors, with much of the decision-shaping happening through self-guided content and internal sharing. The broader takeaway isn’t that buyers behave fundamentally differently by country, it’s that modern B2B buying is globally digital-first, committee-driven, and self-directed. The organisations that succeed internationally are the ones that design demos not just for the person in the meeting, but for the entire invisible buying group evaluating in parallel.

Sławomir Nitek, Angel Investor, Member of Vestbee Angels

Vestbee Angels’ member Sławomir Nitek observes how Polish startups increasingly compete globally through strong technology and deep customer understanding, particularly in AI-driven SaaS for sales. He emphasizes that the most valuable companies today solve concrete business problems while designing for international scale from the outset, noting Demoboost’s enterprise focus, AI and analytics capabilities, product-first mindset, and commercially aware team. As Mr Nitek points out, demo automation and data-driven sales decision-making address urgent SaaS market needs by improving efficiency and shortening sales cycles, thus making Demoboost an apparent investment decision.

Entering Demo-Led Revenue

Today, Demoboost is entering the next chapter, moving beyond demo automation into what the co-founders call Demo-Led Revenue. The new EUR 2.8M funding will primarily accelerate product development, particularly AI-driven demo creation, personalisation at scale, and building out our revenue intelligence capabilities. The startup is doubling the team by hiring 12 additional people across engineering, customer success, marketing and sales. The company is also strengthening its presence in Europe and expanding further in the US and Canada.

‘Success for us means continuing that trajectory while establishing ourselves not just as a demo automation tool, but as a Revenue Intelligence Platform — going beyond demo creation to unify demo engagement analytics with opportunity behaviour data and predictive revenue signals.

At the heart of that is product acceleration. Thanks to our very close relationships with clients, we’ve developed a clear and validated roadmap. Now it’s about executing faster, especially around AI-driven capabilities, deeper analytics, and expanding our revenue intelligence layer. We’ll continue focusing strongly on Sales and Presales use cases, because that’s where we see the strongest and clearest product-market fit,’ Ms Decroix shares.

‘But scaling isn’t just about building more features. It’s also about doing it responsibly. We’ve built Demoboost on going above and beyond for our clients, and maintaining that level of care and responsiveness as we grow is non-negotiable for me. And just as importantly, culture. We’ve built a highly motivated, collaborative team, and as we expand globally, protecting that culture matters as much as any product or revenue milestone,’ the CEO concludes.

In a B2B landscape where buying happens largely outside the meeting room, Demoboost is reframing the product demo from a static presentation into a dynamic source of strategic intelligence. By capturing real behavioural signals and turning them into actionable revenue insights, the company is addressing one of the most persistent blind spots in SaaS sales: understanding who is truly engaged and why. This shift from demo automation to demo-led revenue is poised to redefine how modern sales teams prioritise effort, allocate resources, and ultimately compete in an AI-driven market.

Demoboost is envisaging a Series A round within the next 12–18 months.

The post Demoboost Attracts EUR 2.8M to Turn B2B Demos into Revenue Intelligence appeared first on ITKeyMedia.

We are witnessing the rapid evolution of autonomous AI agents. They have already moved far beyond the role of “smart chatbots”: they work with external content, files, and APIs, interact with payment services, and can operate with one another without human involvement. A distinct digital ecosystem is emerging—one with its own logic, a new attack surface, and a fundamentally different speed at which risks can spread.

Individual incidents such as Moltbook have already demonstrated possible scenarios involving API key leaks, unauthorized data access, and the hijacking of bots. This is not accidental; the risk is systemic in nature.

Language models interpret instructions and data as plain text, without a clear boundary between “what to do” and “what to work with.” That is precisely why prompt injection remains one of the most serious threats. In its LLM Top 10, OWASP ranks prompt injection as the number one risk for LLM-based applications, as it enables direct manipulation of model behavior without exploiting traditional software vulnerabilities.

If an AI agent has access to files and tools or is able to act within external systems, a single injection can quickly escalate from “strange behavior” to unwanted operations and full environment compromise. Even multi-layered defenses degrade when agents are granted excessive privileges and lack clearly defined execution boundaries. Relying on a “quick patch” is therefore unrealistic—what is required are foundational architectural constraints, segmentation, and action validation by default.

The interaction between agents introduces additional complexity. Malicious instructions are transmitted as ordinary messages, tool substitution redirects requests to fake endpoints, and injections propagate through integrated data sources. Under such conditions, isolated failures can easily turn into cascading incidents.

Are we ready? Only partially. For example, the AI Risk Management Framework by NIST offers a high-level foundation for addressing risk. In practice, however, such principles as minimized privilege, trust segmentation, environment isolation, and “human-in-the-loop” controls often turn up only after the first incidents occur.

Risk increases sharply when agents begin exploiting each other’s weaknesses faster than security teams can detect and respond. For that reason, strict privilege limitation, control over sources and tools before execution, continuous behavioral monitoring, and a rethinking of architecture for environments with unpredictable agent interaction must become priorities.

The post The Security Challenge of Autonomous AI by Olga Voloshyna appeared first on ITKeyMedia.

• Estonian-Ukrainian defense tech startup C2GRID closed Pre-Seed round led by Sunfish Partners
• C2GRID’s platform transforms drone footage into near-real-time 3D battlefield environments
• The solution is specifically designed for GPS-denied, security-sensitive, frontline operational conditions
• The startup’s future expansion targets AI-driven planning, simulation, and autonomous integration

Last November, C2GRID—the Estonian-Ukrainian 3D geospatial situational-awareness platform for the military—closed its Pre-Seed round of investment. The lead investor of the round was the famous VC firm Sunfish Partners (invested in CodeAlly, among others), joined by another well-known investor BSV Ventures (invested in FirstQFM, among others) and Mike Oliinyk of Montonio as an angel investor.

From Research Environment to Real-World Urgency

C2GRID was incorporated in 2024 but the team was formed earlier at TalTech’s Industrial Virtual and Augmented Reality Lab, led by Professor of the Department of Mechanical and Industrial Engineering, Vladimir Kuts.

This lab focuses on systems modelling and autonomous platforms. Having worked closely with students and researchers in the lab environment for years, the core team naturally formed around people with strong technical backgrounds and the willingness to take this forward. Since several lab members are from Ukraine, they were motivated to apply their skills in a situation where their homeland is facing Russian aggression.

Andri Sokka, Technology Transfer Expert at TalTech

‘Our team wanted to do something with the potential to save lives in Ukraine,’ C2GRID’s CMO Saamuel Stepanov tells ITKeyMedia.

‘One of TalTech’s goals is to encourage and support employees’ entrepreneurial initiatives, as spin-off companies are a clear example of the impact of the university’s research and knowledge on society. We are very pleased that C2GRID has grown out of TalTech and has managed to secure an investment that creates the conditions for faster growth and market success,’ TalTech’s technology transfer expert Andri Sokka adds.

Frontline Validation and Early Momentum

The first major milestone for the company was direct engagement with Ukrainian frontline units. C2GRID attended Darkstar’s hackathon to test the idea with well-known units. They responded positively, and that gave the team the confirmation to pursue what they were building.

Following that validation, C2GRID formalized the company and secured initial funding in September 2025. This allowed the startup to move from early experimentation to structured product development and begin expanding its operations in Ukraine.

Since then, the young company has widened its engagement with Ukrainian units – deepening cooperation, collecting field feedback, and adapting the solution to real operational constraints. In parallel, C2GRID has engaged Estonian and NATO structures, as well as several defence primes, to ensure interoperability and long-term integration potential.

Getting selected into NATO DIANA marked another important milestone. It provided recognition within the Alliance innovation ecosystem and connected C2GRID to a broader international network of partners and stakeholders.

Turning Raw Footage into Spatial Intelligence

Vladimir Kuts, CEO at C2GRID

‘Opportunities in 3D mapping for defence and civilian use are expanding rapidly, with new ones emerging almost daily and no sign of slowing, quite the opposite. With massive amounts of video generated on today’s battlefields, our near real-time 3D solution handles virtually any footage, giving those on the front lines better tools to defend freedom – today or whenever necessary,’ Mr Kuts states.

Essentially, C2GRID is a deep-tech geospatial awareness platform that transforms raw battlefield video, particularly drone feeds, into accurate, near-real-time 3D operational environments to support situational awareness and decision-making for defence and security users. It fuses multiple sensor data sources and layers, applies AI-based automation (including threat detection), and creates unified, multilayered maps that help analysts and commanders plan missions and respond faster in complex environments. Being co-developed with real frontline units, the system is designed to be hardware-agnostic and usable in contested settings without reliance on GPS.

‘What C2GRID is building is the next step on the battlefield. Instantly duplicating the environment into a 3D virtual space that benefits from data fusion opens an entirely new layer of possibilities — from better situational understanding to faster planning, training, and more efficient use of unmanned and autonomous systems,’ Mr Kuts summarizes.

Engineering for Imperfect Conditions

As with any computer vision and 3D reconstruction system, input quality influences output fidelity and higher resolution, stable footage, and clear visual references naturally improve model accuracy. Operational environments, however, introduce additional complexity. Weather conditions, lighting variability, terrain characteristics, and signal constraints can all affect data capture and processing. On the battlefield, speed and efficient data flow are often as critical as raw image quality.

‘Our focus is on optimising performance under real-world conditions rather than ideal ones. We continuously refine our algorithms, processing pipelines, and system architecture to maximise usability across diverse scenarios. Close cooperation with operational partners ensures that feedback is rapidly incorporated into product development,’ Mr Stepanov argues.

Navigating Without GPS

Saamuel Stepanov, CMO at C2GRID

Maintaining spatial accuracy in GPS-denied or GPS-spoofed environments requires alternative referencing methods. Rather than relying solely on satellite positioning, C2GRID primarily uses vision-based spatial reconstruction and relative positioning derived from the video itself.

The system can anchor reconstructed models to geographic reference points or pre-existing maps when available. In operational scenarios, approximate coordinates can be introduced to align the model within a broader geospatial framework, with users able to fine-tune positioning if required.

‘The tradeoff is that absolute global accuracy may vary depending on available reference data and environmental conditions. However, for many operational use cases, relative spatial accuracy and structural coherence are more critical than centimetre-level global precision. Our focus is on delivering reliable, actionable spatial awareness under conditions where GPS cannot be used,’ Mr Stepanov tells ITKeyMedia.

Security as Architecture

Given the sensitivity of the data involved, including video from frontline operations, potentially featuring civilians, C2GRID needs to address security, data protection, encryption, and the risk of data falling into the wrong hands. According to the team, security and data protection are foundational design principles for C2GRID. From the outset, the platform has been architected with secure deployment models in mind, including on-premise configurations to ensure full control over sensitive operational data.

The platform’s development is aligned with relevant NATO security and interoperability standards and closely monitored evolving compliance requirements within allied defence ecosystems. Participation in exercises such as CWIX is part of C2GRID’s roadmap to validate interoperability and security in real operational environments.

‘Given the sensitivity of frontline video, including potential civilian exposure, encryption, access control, and controlled data handling are treated as core system requirements, not add-ons. This is an area that requires continuous attention, and we have prioritised security from day one of development,’ Mr Stepanov assures.

Human Judgment in the Loop

Another challenge is to ensure that commanders, analysts, or on-the-ground units trust the AI-driven output and effectively interpret it. De facto this presupposes market education under extreme stress. Seeing how critical trust is especially when decisions are made under stress, C2GRID’s approach is to position its AI as a supportive tool rather than a possible replacement for human judgment.

The system highlights detected changes and objects in the areas of interest, ensuring that outputs remain transparent and traceable to the underlying data. Analysts and commanders retain full control over interpretation and decision-making, with the ability to verify findings directly within the reconstructed environment. Human oversight is embedded into the workflow by design, and the platform is built to augment situational awareness, not to automate critical decisions without supervision.

‘As capabilities evolve, we continue to evaluate explainability, user interface clarity, and training requirements to ensure outputs remain interpretable and trusted in real-world conditions,’ Mr Stepanov comments.

From Validated Prototype to Deployable Product

Max Moldenhauer, Founding Partner at Sunfish Partners

With the new funding, C2GRID can focus on maturing the product and transitioning from a validated prototype to a robust, commercially deployable product. This includes hardening the system architecture, improving performance and stability under operational constraints, refining the user interface based on frontline feedback, and standardising deployment workflows. The goal for Q1 2026 is an operationally validated, deployable, and commercially ready product.

‘C2GRID’s ability to turn raw battlefield data into actionable intelligence in near real time is not just a breakthrough for defence – it is a transformative step for how humans interact with complex environments. The team strengthens European security and pioneers applications that could one day extend far beyond the battlefield,’ Sunfish Partners’ partner Max Moldenhauer firmly believes.

‘For today’s defence operators, raw video alone doesn’t cut it — situational awareness is what they really need. That depends on rapidly turning footage into something actionable. C2GRID’s approach directly addresses this. Vladimir and his team combine both deep technical capability and real operational understanding, which is crucial to solving problems in the field. They’ve already proven results on the front lines and are building on that momentum. This team can deliver, and we’re committed to enabling their growth in serving European and NATO-allied defence and security goals,’ BSV Ventures’ general partner Erik Bhullar adds.

Beyond 3D: The Next Layer of Capability

Erik Bhullar, General Partner at BSV Ventures

While 3D reconstruction and situational awareness form the foundation of the platform, our longer-term direction includes expanding AI-supported capabilities across the operational workflow. C2GRID envisions extending beyond 3D reconstruction and situational awareness and into automated mission-planning, predictive simulation, or even coordination of autonomous/unmanned systems. 

This includes deeper analytical tools, enhanced change detection, automated pattern recognition, and decision-support features that build on the spatial data layer that C2GRID is already generating. The focus is on augmenting planning, coordination, and information flow.

In parallel, C2GRID is developing a streaming and integration layer designed to unify video feeds, spatial models, and additional sensor inputs into a single operational environment. The goal is to create a cohesive platform that supports analysis, planning, and coordination without fragmenting workflows across multiple systems.

Andrius Milinavičius, General Partner at BSV Ventures

Future extensions may include tighter integration with autonomous and unmanned systems, but these developments can only be driven by operational demand and validated use-cases.

‘Vladimir and his C2GRID team combine deep technical capability and real operational understanding which is crucial to solving problems in the field. They’ve already proven results on the front lines, and are building on that momentum. This team can deliver, and we’re committed to enabling their growth in serving European and NATO-allied defence and security goals,’ BSV Ventures’ general partner Andrius Milinavičius concludes.

C2GRID sits at a peculiar and pivotal intersection of computer vision, geospatial science, and the brutal realities of modern warfare, where information speed can mean the difference between chaos and coordination. By transforming raw, fragmented battlefield video into coherent, navigable 3D environments, the company is reshaping how decisions are made under pressure, where clarity is scarce and time is unforgiving. Succeeding in turning this from a promising prototype into a trusted operational standard will define C2GRID’s new layer of digital infrastructure for defence, where data is not merely collected, but truly understood.

The post C2GRID Raises Pre-Seed to Reinvent Situational Awareness appeared first on ITKeyMedia.