LeadDesk joins forces with AMD and QuTwo to build a vertically integrated, sovereign AI stack designed to improve conversational intelligence for Nordic languages and regulated European sectors.
LeadDesk is integrating QuTwo's quantum-inspired AI compression and AMD's compute infrastructure to develop specialized conversational AI models for Nordic languages. This initiative aims to provide a competitive advantage by offering real-time agent guidance and voice automation that outperforms general-purpose models in low-resource European languages. By embedding these capabilities directly into its platform, LeadDesk seeks to strengthen its position in regulated sectors like healthcare and finance.
To build a vertically integrated AI stack that provides optimized, sovereign conversational AI models for European languages, starting with Nordic languages.
It enables real-time agent guidance, natural voice AI in European languages, and trusted services for regulated sectors like public services and finance.
QuTwo is positioning itself as a key provider of enterprise-grade AI by supplying its quantum-inspired model compression technology to the LeadDesk platform. This partnership validates QuTwo's 'intelligence layer' strategy, demonstrating how its research can be applied to real-world, business-critical customer interactions. The collaboration serves as a practical deployment of its AI stack, moving from research to industrial application.
QuTwo provides the AI research and quantum-inspired model compression that allows the AI models to run efficiently at scale within the LeadDesk platform.
QuTwo focuses on building AI for the quantum era, creating an intelligence layer for industrial AI that supports classical, hybrid, and full quantum AI systems.
AMD provides Instinct GPUs, EPYC CPUs, and ROCm software for the sovereign AI stack, reinforcing its position in European AI infrastructure.
LeadDesk Plc has announced a strategic collaboration with AMD and QuTwo to develop optimized conversational AI models specifically for Nordic languages. By integrating AMD's high-performance hardware with QuTwo's quantum-inspired compression, the partnership aims to deliver sovereign, efficient AI solutions that outperform general-purpose models in regional customer service applications.
Most frontier AI models are trained primarily on English, leading to performance gaps in smaller European languages regarding speech recognition, cultural context, and the understanding of local names and numbers.
AMD provides the foundational hardware and software infrastructure, including Instinct GPUs, EPYC server CPUs, and the ROCm open software platform, to power the efficient execution of these specialized AI models.