Multiverse Computing

🚀 We're growing, and we're hiring. Multiverse Computing has +15 open positions across Europe and the US. If you want to build efficient, secure AI models with a team that ships, this is the place. Find your next challenge here: https://lnkd.in/g2R8c6K9

🛡️ Multiverse Computing and FIOR Group are partnering to deliver secure edge and sovereign AI infrastructure for enterprises, governments and regulated industries. We combine FIOR's cryptographic agent identity and governance technology with Multiverse Computing's efficient AI models and edge deployment capabilities, so organizations can run AI inside their own infrastructure without giving up control over identity, access, policy compliance and data sovereignty. 🔑 FIOR's agent identity gateway, policy enforcement, tamper-evident audit trail, mobile protection and voice authentication integrate into Multiverse Computing-enabled AI deployments. ⚙️ Multiverse Computing supports the deployment and optimization of AI models within FIOR-enabled environments, including in-perimeter execution and anomaly detection. The integrated solution targets high-assurance environments: sovereign AI infrastructure, regulated enterprise AI, defense and public-sector deployments, telecom and edge AI, industrial automation, financial services and healthcare. 🔗 Learn more: https://lnkd.in/gxzUPj7F #AI #SovereignAI #EdgeAI #AIInfrastructure #AIGovernance #MultiverseComputing #FIOR

🧠 Most AI projects don't stall on ambition: they stall on compute, memory and infrastructure costs. That is the conversation Mark Rabkin had with Intel Business, and our answer to it starts on the model side. Multiverse Computing builds its own ultra-efficient AI models, optimized with CompactifAI, so performance stays high and the footprint drops. As an HPE and Intel partner, we run those models on hardware built for AI. The result is not just a cheaper deployment, it is AI that becomes viable in environments where it previously wasn't. #AI #EfficientAI #EnterpriseAI #CompactifAI #Intel #HPE

⛵️ Podium at the Copa del Rey Mapfre. The GP42 GARM, sponsored by Multiverse Computing through lead investor GP Bullhound, finished third in Sail Racing ORC A at the 44th Copa del Rey Mapfre in Palma de Mallorca, one of the most competitive fields the class has seen in years. Skippered by Per Roman, founder of GP Bullhound, GARM went into the final day fighting for the top spots after a sixth-place finish at the 2026 ORC World Championship. The class title came down to a one-point duel between Nola (Estonia) and Rán (Sweden), with Nola taking the win and the overall Copa del Rey Mapfre title. GARM held on to complete the podium. Trophies were presented by King Felipe VI at the Palacio de la Almudaina, closing out a week that turned the Bay of Palma into the center of international sailing. Congratulations to Per Roman and the entire GARM crew on a season, and a Copa del Rey, to remember. More info here: https://lnkd.in/eCZznKT7 #MultiverseComputing #CopaDelReyMapfre #Sailing #GPBullhound

Is your AI safe once it's actually running, not just before you deploy it? 🛡️ In his latest Forbes Technology Council article, our CTO and co-founder Samuel Mugel, PhD explores why enterprise AI's governance gap is real, and why runtime safety is the missing layer. Pre-deployment testing catches what red-teaming anticipates. It doesn't catch what happens once a live workflow meets real users, adversarial inputs and other AI agents acting on its behalf. At Multiverse Computing, we build AI with efficiency and security as first principles. As enterprises move from single models to multi-agent workflows and RAG pipelines, governance has to move with them: watching retrieval decisions, inter-agent communication and tool calls in real time, not just the final response. Read Sam's full article here 👇 https://lnkd.in/et_Yqnw5 #AI #EnterpriseAI #AIGovernance #RuntimeSafety #ResponsibleAI #MultiverseComputing

We are pleased to welcome Qatar Development Bank (QDB) as an investor in Multiverse Computing and announce that Doha will become our regional headquarters 🇶🇦 QDB’s investment forms part of our €500 million Series C funding round. From Doha, we will serve customers and develop partnerships across the Middle East, bringing our efficient and sovereign AI technologies closer to one of the world’s fastest-growing innovation ecosystems 🚀 Our AI models and solutions enable more secure, cost-effective and energy-efficient AI deployments across cloud, data center and edge environments ⚡️ We look forward to working with QDB and the Startup Qatar Investment Program to support knowledge exchange, regional growth and Qatar’s ambitions in sovereign AI. 🔗 Learn more: https://ow.ly/PbBe50ZwSCO #MultiverseComputing #QatarDevelopmentBank #StartupQatar #EfficientAI #SovereignAI #Doha #ArtificialIntelligence

🤝 Multiverse Computing and Qualcomm Technologies are collaborating to bring more efficient AI models to data centers worldwide. Our AI models will be specialized for Qualcomm Dragonfly™ AI200 and AI250 accelerators, combining Qualcomm Technologies’ AI acceleration hardware with our model compression and optimization technology. For data center operators, this means greater capacity from the same infrastructure: more inference requests, more models running concurrently, and the ability to scale AI services with lower compute, memory, and energy requirements. The potential was demonstrated at Mobile World Congress 2026 using the Qualcomm® Cloud AI100 Ultra accelerator: ⚡ Up to 93% faster response times 📈 Up to 54% higher throughput 💾 Up to 45% lower memory usage 🌱 Up to 21% lower power consumption 🎯 No loss in accuracy Together, we are creating a practical path toward scalable, high-performance, and energy-efficient AI infrastructure. 🔗 Learn more: https://ow.ly/UyoO50Zwrcq #MultiverseComputing #Qualcomm #EfficientAI #DataCenters #AIInfrastructure #ModelOptimization

🤝 Proud to collaborate with Plain Concepts Research on a new n8n integration for CompactifAI. This node makes it easier for organizations to bring optimized AI models into no-code and low-code workflows, helping turn AI experiments into efficient, scalable, production-ready solutions. Thank you to the Plain Concepts team for bringing this integration to life! #CompactifAI #n8n #AI #Automation #EnterpriseAI

🛡️ Introducing SentinelAI: Multiverse Computing’s new control layer for enterprise AI security and governance. As AI becomes embedded across enterprise workflows, organizations need greater visibility into what data is sent to models, how it is processed and whether each interaction complies with their policies. SentinelAI is designed to provide that control in real time. Deployed within an organization’s own infrastructure, SentinelAI sits between users, applications, agents and models. It enables enterprises to: 🔒 Detect sensitive and confidential information 🚫 Allow, block, redact or flag interactions based on custom policies 🧱 Identify prompt-injection and jailbreak attempts 💰 Monitor token usage and uncontrolled AI spending 📋 Create structured, audit-ready records of AI activity SentinelAI is Kubernetes-native and model-agnostic, providing consistent governance across commercial, open-source, in-house and Multiverse compressed models without requiring organizations to rewrite their existing applications. Its governance engine uses lightweight, specialized Multiverse models to deliver contextual risk detection, while the governance infrastructure, detection models and audit records remain within the customer’s environment. 🔗 Learn more: https://lnkd.in/eZPNvWZE #MultiverseComputing #SentinelAI #AIGovernance #AISecurity #EnterpriseAI #DataSovereignty #ResponsibleAI

This is exactly the kind of ecosystem innovation that accelerates AI adoption across Europe and beyond. Great to see Intel and Multiverse Computing showing what's possible when advanced model compression meets Intel® Xeon® 6 CPU. The result: efficient AI inference on existing CPU infrastructure, helping organizations deploy AI faster, more cost-effectively, and at scale. #AI #EnterpriseComputing #Xeon #Innovation #TechLeadership https://lnkd.in/dtWpVcKi