Led · Seed · Apr 2026

Cnuic has developed a working prototype of a completely new type of photolithography device that leverages properties of light to enable rapid, reconfigurable production of photonic chips with enhanced 3D control. The company says this approach allows manufacturing possibilities previously impossible and could unlock mass production of photonic components. Cnuic positions its technology as a potential enabler for higher-speed, lower-power data-centre hardware and accelerated AI model training by removing communication bottlenecks among processors. The company also highlights broader applications for its technology across metalenses, 3D photonic crystals, AR/VR waveguides, flexible gratings, and other light-based systems. Cnuic is based in Edinburgh and is pursuing commercialization and scaling of its prototype following the pre-seed raise. Investors describe the technology as a major deep-tech breakthrough that could strengthen Europe’s role in the global semiconductor industry.

Participated · Seed · Apr 2026

GRAI develops consumer-facing AI music products designed to make music interaction social and participatory rather than focused solely on generative creation. Its current products include the iOS remixing app Music with Friends and an AI music playground for Android, which the company uses to gather consumer feedback. GRAI has built its own taste and participation graph and is developing a derivatives pipeline and real-time audio systems to preserve original-track identity while allowing transformations. The founders are Belarusian and previously sold the video app VOCHI to Pinterest; the CEO Ilya Liasun is based in Poland. GRAI aims its products at Gen Z and Gen Alpha users and is working with artists and labels to create an opt-in framework for remixing that could generate new royalty streams. Financially, the company has secured $9 million in seed funding to advance its product roadmap and infrastructure.

Participated · Seed · Apr 2026

Edmund builds AI agents that integrate technical documentation, PLC projects, maintenance logs and real-time machine data into a single operational layer for factories. Its platform guides technicians to identify faults, determine root causes and receive step-by-step troubleshooting, cutting diagnosis time by up to 90 percent. In customer deployments, such as at Amcor Flexibles, Edmund reduced average repair times by 26 percent, saving roughly 440 man-hours per factory annually. Founded in 2023 and designed to be hardware-agnostic, the company plans to use its recent €2.5M raise to expand its team and enter additional European and US markets while further developing its contextual AI troubleshooting capabilities.

Participated · Seed · Feb 2026

Prague-based ValkaAI is building a platform for fully controllable digital humans that can react to users in real time, turning traditionally passive video into an interactive medium. Founded by Vlastimil Venclík and Miloš Lokajíček and operating under Realms Group, the startup focuses first on AI commentators that analyze live sports or esports action and speak to each viewer in their preferred language. Its technology addresses common challenges in generative video such as latency, visual flicker and high compute cost, enabling scalable, personalized interactions. The company employs roughly 45 people across San Francisco, Prague and Dubai, including researchers from Meta and DeepMind. Fresh capital will fund expansion of research and product teams in the United States, Europe and the UAE, strengthen the core platform and scale computational infrastructure. Near-term goals include launching commercial pilot integrations with distribution partners in sports and esports. The recent financing is the company’s first external raise and positions it to build a globally relevant entertainment technology stack.

Led · Seed · Feb 2026

AppFactor delivers an agentic orchestration platform that automates the ongoing maintenance and modernization of complex enterprise applications. Operating through a network of specialized AI agents, the system continuously discovers running workloads, pinpoints vulnerabilities, performs bug fixes, and executes dependency updates without human intervention. Its dynamic discovery engine maps the entire application topology—repositories, dependencies, infrastructure, and runtime behavior—so autonomous actions remain fully contextualized and auditable via standard pull-request workflows. A standout capability is autonomous code regeneration into Rust, enabling organizations to migrate legacy COBOL or compute-heavy workloads for improved performance, lower costs, and stronger security. Large enterprises are already using the product to shrink modernization timelines and adopt cloud-native patterns such as serverless compute and managed databases. The company targets the estimated $1.5 trillion technical-debt backlog across sectors like financial services, healthcare, retail, and manufacturing. Fresh funding will deepen the platform’s autonomous refactoring features and fuel go-to-market expansion. The announcement did not disclose any revenue, user, or profitability metrics.