Leapcraft orchestrates healthy and efficient buildings. We work with international clients from the public and private sector to deliver actionable insights into the climate in and around their buildings to help them optimize for energy savings and better productivity. We are thought leaders in IoT-sensors and data infrastructure for environmental monitoring.

Pangea Accelerator is an investment and competence development platform that connects talented and impact-driven start-ups with international angel investors and African diaspora investors. We won the Nordic Impact Award 2018 at Copenhagen Fintech & partner with UN-habitat, Microsoft and many more.We provide three months milestone based accelerator program focusing on validation & scalability. We also provide investor training programs in Oslo and StockholmWe are based in Oslo and Nairobi :)

At Sana Lab we're a team of scientists and engineers who have worked at places like Apple and NASA developing a personalisation algorithm for education. Our in-house research makes innovative use of recurrent neural networks in order to bring personalised learning to the world.

Wolt is the one app for great food. Over 1000 restaurants in Finland, Sweden, Denmark, Estonia, Latvia & Lithuania. Discovery, takeaway, home delivery.Whether you have a craving for your favourite meal from the restaurant around the corner, or you’re hungry but not sure what you want to eat, Wolt will help you find it.

Founded in 2017, Habity is a Stockholm-based technology company passionate about simplifying your housing. From the overall market to the services and everything in between.

Costile is an AI agent diagnostic proxy for teams building production AI workflows. As companies start using AI agents for support, research, coding, internal operations, and automation, they run into a new kind of infrastructure problem: the agent may technically work, but nobody knows when it is wasting money, looping, retrying badly, using the wrong model, or silently drifting outside expected cost behavior. Most teams can see their total AI bill after the fact. That is too late. What they need is operational visibility into which agent caused the spend, what pattern caused it, how severe it is, and what action should be taken. Costile sits between a company’s AI agents and model providers such as Anthropic. It routes requests through a proxy, logs only metadata such as model, tokens, cost, timestamps, agent ID, session ID, and stop reason, and turns that metadata into diagnostics. Prompts and responses are not stored. The product shows teams a command-center view of their AI agents: which agents are healthy, which are critical, what they spent in the selected time window, and what issue is driving the risk. When an agent is flagged, Costile gives a diagnostic report with the incident window, request log, baseline comparison, projected monthly exposure, and recommended fix. The problem we are solving is not just “AI is expensive.” The deeper problem is that AI agents are becoming autonomous software systems, but companies are still monitoring them like simple API calls. A bad agent loop, retry cascade, model-routing mistake, or runaway request can create cost, reliability, and governance risk before anyone notices. Costile helps engineering, product, and operations teams answer three questions quickly: What is going wrong? Which agent or session caused it? What should we do now? The initial product focuses on AI cost diagnostics, haywire detection, alerts, audit logs, API key management, and enterprise-ready reporting. Over time, Costile can become the observability and control layer for AI agent operations.

Cobrief is developing an AI-powered public procurement platform to help businesses across all industries find, evaluate, and win government tenders — replacing manual, time-consuming bid processes with intelligent automation that acts like having an extra colleague on the team. Public procurement represents over 800 billion NOK annually in Norway alone, yet the average government tender receives only around 3–5 bids. In nearly 20% of tenders, only one or zero suppliers submit a bid. The reason: tender data is scattered across closed platforms, the process is complex, and small businesses lack the resources to keep up. Suppliers can't find the opportunities, or give up because the process is too complicated. The result is less competition, worse deals for taxpayers, and lost revenue for businesses that could have won. Cobrief aggregates public tenders from procurement databases across Scandinavia, Finland, and Germany — including Doffin, TED, Udbud, and Hilma — into a single searchable platform. AI-powered matching scores every opportunity against each company's profile, sending relevant alerts via email, Slack, or Teams. When a team decides to bid, an AI writing module drafts proposal responses from a knowledge base of past submissions. Automated price sheet filling pulls from product catalogs, and CV matching identifies the right personnel. A Bid Manager Agent provides autonomous support throughout the entire process — functioning as an AI colleague from evaluation to submission. Teams collaborate in shared workspaces with task boards, deadline calendars, and role-based access. Used by 1,500+ companies, with a free plan to get started. The founders experienced firsthand how broken the public procurement process is for suppliers. Tender data that should be freely and directly accessible by law is in practice locked behind commercial platforms, creating unnecessary barriers to competition. They saw that businesses — especially SMBs — were missing out on opportunities worth billions simply because the system was too opaque and complex. Cobrief was built to fix that: make public tenders accessible, understandable, and winnable for every business. Beyond the product, the team is actively advocating for open procurement data in Norway through anbudsdata.no, pushing for systemic change so that public tender data is truly open — not just in theory, but in practice.

Based in Norway, Recharge Health is a team of innovators dedicated to unlocking the full potential of the human body. Through collaboration with doctors and researchers worldwide, we've created FlexBeam, a revolutionary recovery device utilizing targeted infrared light therapy. Our mission is to bring the power of infrared light therapy into every home. More on FlexBeam: FlexBeam is the most powerful infrared therapy device of its kind. This non-pharmaceutical, non-invasive solution is already making a significant impact on the lives of many. With over 40,000 satisfied customers worldwide and endorsements from notable, global athletes, this wearable recovery device is changing the game of athletic recovery and health optimization. Developed with NASA-inspired technology, FlexBeam uses infrared light therapy to deliver powerful wavelengths of red (610-650nm) and near-infrared (800-850nm) light at up to 110mW/cm2, and a total optical power output of 5.5W. This powerful and targeted light stimulates your cells to produce more energy – so you can repair muscle, relieve pain, and recover faster. Learn more about our company here: https://www.youtube.com/watch?v=iEv3BimOCS8

Most companies know they need to move on AI. The challenge isn’t awareness - it’s knowing where and how to start and keep momentum. That’s the kind of problems we work on at Kompose. Working at Kompose means being part of a small, highly focused team that builds side by side. We combine deep technical expertise with hands-on delivery, helping professional services firms and larger enterprises move from complexity to working, production-ready AI solutions. No slide decks for the sake of it. No pseudo-work. We focus on real use cases and ship solutions that deliver visible results. We move fast and prioritise momentum. If something doesn’t create value for the customer, we don’t do it. Good ideas win - regardless of title. Feedback is direct and constructive. Mistakes are fine; carelessness and unnecessary complexity are not. Our days are collaborative. We start each morning with a short check-in and spend most of our time working closely together in the office in Christianshavn. There’s space for deep focus and space for discussion. Transparency is the norm - about challenges, trade-offs, and decisions. We experiment constantly and apply AI in our own workflows. We test new tools, refine how we build, and continuously evolve our delivery framework so we get sharper with every project. You’re trusted to own your work and see it through - from idea to production. We’re ambitious and growing organically, but we believe good work requires balance. Some weeks are intense, others lighter. We swim in the canals every week (even in winter) and during summer we hold meetings on a rented boat. Lean team. High trust. Real responsibility. Real impact.

ExecuteAI is building a unified AI‑native ecosystem designed to make advanced automation, agent development, and AI‑driven operations accessible to every company—regardless of size, technical maturity, or industry. We believe AI should not be limited to large enterprises with dedicated engineering teams. Instead, it should be a practical, reliable, and affordable tool that any organization can deploy to improve productivity, decision‑making, and customer experience. Our mission is to democratize AI by providing a complete operating system for automation: from outreach and CRM orchestration to agent development, workflow execution, and talent collaboration. At the core of our platform is the ExecuteAI Hub, a login‑based workspace where companies and AI developers meet, collaborate, and deliver real solutions. The Hub functions as a marketplace, a project management environment, and a delivery engine in one. Companies can post AI projects, describe their needs, and connect with qualified AI developers who specialize in automation, agent design, data workflows, and integration engineering. Developers gain a structured environment to showcase their skills, build a portfolio, and deliver high‑quality AI solutions with clear expectations, transparent communication, and standardized processes. We are solving a fundamental problem in the AI industry: fragmentation. Today, companies must navigate a confusing landscape of tools, models, consultants, and platforms. There is no single place where they can reliably define a problem, find the right expertise, and receive a complete, production‑ready AI solution. ExecuteAI eliminates this friction by offering an end‑to‑end ecosystem that covers the entire lifecycle—from idea to deployed agent. Our platform includes tools for outreach automation, CRM orchestration, agent execution, workflow design, and continuous improvement. Everything is built with a strong focus on Nordic values: transparency, trust, quality, and responsible AI governance. For companies, ExecuteAI provides clarity and structure. Instead of hiring expensive consultants or experimenting with disconnected tools, they gain access to a curated talent pool and a standardized delivery framework. For developers, we offer a place to work on meaningful AI projects, collaborate with clients, and build long‑term relationships. For both sides, the Hub ensures that AI development becomes predictable, repeatable, and scalable. ExecuteAI is headquartered in Copenhagen and built on the belief that the future of work will be shaped by human‑AI collaboration. Our goal is to empower organizations to operate faster, smarter, and more efficiently by integrating AI into their daily workflows. Whether a company needs a custom agent, a full automation pipeline, or ongoing AI operations, ExecuteAI provides the tools, talent, and infrastructure to make it happen.

Defend Denmark er Danmarks første og eneste lokale bug bounty platform. Vi forbinder danske etiske hackere med danske virksomheder, så sårbarheder bliver opdaget og lukket inden de udvikler sig til reelle trusler. På platformen kan etiske hackere lede efter sikkerhedshuller hos vores kunder. Finder en hacker en reel sårbarhed, udbetaler kunden en dusør. Det gør it-sikkerhed resultatorienteret frem for tidsorienteret, som ellers er normen i konsulentverdenen. Målet er at få flere danske virksomheder til at tage bug bounty modellen til sig som et aktivt værktøj i deres sikkerhedsarbejde. Defend Denmark er søstervirksomhed til islandske Defend Iceland. Vi er stadig tidligt i opbygningen af virksomheden og kommende medarbejdere får reel indflydelse på de praktiske rammer, herunder kontorplacering og daglige arbejdsgange.

Velocity Insights is building a new class of analytical production systems that generate decision intelligence from a fully traceable evidence base. Most commonly used GenAI approaches generate narratives first and attempt to justify them during the text generation process — often resulting in outputs that are incomplete, difficult to verify, and not fully traceable to underlying evidence. Velocity Insights takes a different approach, structuring the evidence first and deriving every insight from it. This enables organizations to produce analytical outputs that are fully traceable, inspectable, and auditable — transforming how conclusions are validated in high-stakes decision environments. Designed for capital allocation and strategic decision-making, Velocity Insights enables enterprises across finance, life sciences, and technology to generate high-speed decision intelligence without sacrificing traceability or control. By combining data structuring, workflow design, and AI, the system establishes a new standard: analytical outputs that are fully verifiable, transparent, and trusted.

maiaa is a curated marketplace for design-led meeting and event spaces, helping companies and teams find venues with character for meaningful, in-person gatherings. We combine carefully selected spaces with a thoughtful booking flow to replace generic, transactional venue sourcing with a more human, design-driven experience. The problem we’re solving: Finding inspiring spaces for meetings and small events is still fragmented and inefficient. Companies often rely on a mix of spreadsheets, emails, and generic listing platforms that prioritize volume over quality. The result is wasted time, inconsistent experiences, and venues that don’t reflect the purpose or culture of the gathering. This affects: • companies that value design, brand, and experience • venue owners with distinctive spaces who lack the right distribution and tools Despite a clear demand for better spaces, the market remains operationally outdated and aesthetically undifferentiated. How we solve it: maiaa brings structure, curation, and clarity to the process. We: • curate a limited selection of character-rich venues rather than listing everything • focus on storytelling, aesthetics, and context — not just capacity and price • provide a request-based booking flow that respects how venues actually operate • integrate experiences and add-ons to make gatherings more intentional and complete The platform is being built to balance design sensibility, operational realism, and scalable technology — with a strong focus on usability for both guests and hosts. Why we started: maiaa grew out of firsthand experience. As a founder, operator, and boutique hospitality owner, I’ve spent years working with spaces, guests, and teams — and repeatedly experienced how broken and uninspiring venue sourcing can be, especially for gatherings that actually matter. The motivation behind maiaa is simple: to create a better alternative for people who care about where they meet, how they gather, and the experience they create together. This is not about building another marketplace — it’s about raising the standard for how physical spaces are discovered and used.

AUTOLYZER is transforming the way vehicle damage is assessed. Powered by AI, computer vision, and machine learning, our platform automates the entire damage inspection process—making it faster, more accurate, and fully transparent. We reduce manual effort, eliminate subjective estimates, minimize fraud risk, and help repair shops and insurers operate with greater efficiency and trust.

ABOUT Insurance is one of the biggest recurring household costs - yet the market is built on complexity: dense policies, hard comparisons, and increasingly individualised pricing with little transparency. The result is predictable: people overpay for years, miss critical coverage, or both - and often only find out when it's too late. We're building the consumer layer for private insurance, making it behave like a modern product. Upload your policies and, in under a minute, our AI turns fine print into clear answers: where you're over- or underinsured, where the gaps are, and where you're likely paying too much. We benchmark each policy against an anonymised pool of real policies to recommend better-matched alternatives - shifting knowledge and pricing power back to households. We're early-stage, moving fast, and already proving the need: 50+ customer interviews, 10 households switched, ~28% average savings while improving coverage - plus validation from senior industry professionals, a top-performing insurance salesperson on the team, and initial pre-seed backing from a Bain & Company partner.