The integrated platform combines Multiverse’s CompactifAI compression technology with 1stProtect’s real-time runtime enforcement. CompactifAI shrinks large language models by up to 95% with minimal precision loss, allowing sophisticated AI to function on hardware ranging from industrial edge devices to smartphones. By eliminating the need for cloud-based inference, the solution reduces costs by 50–80% and accelerates processing speeds by up to 12 times.
1stProtect and Multiverse Team Up for Secure On-Device AI
Enterprises can now run powerful generative AI models entirely on-device without cloud connectivity, thanks to a new partnership between Silicon Valley security firm 1stProtect and quantum-inspired compression leader Multiverse Computing. The collaboration aims to solve critical data sovereignty, cost, and runtime security challenges for high-stakes industries.
While compression enables the deployment, 1stProtect’s security layer governs agent behavior, blocking threats such as prompt injection and unauthorized actions within sub-100-millisecond windows. This protection remains functional whether the system is connected to the internet or fully air-gapped. The offering targets sectors where cloud dependency is a liability, including defense, critical infrastructure, healthcare, and financial services, where data privacy and low-latency performance are non-negotiable requirements.
Comments (0)
No comments yet. Be the first!