What if enterprise AI could run entirely on-site? A compact system combines local inference with an AI-powered knowledge base.
ASRock Industrial has introduced AI BOX-A395 bundled with AiKMSpark, an on-premises enterprise AI solution designed to run AI inference and knowledge management locally. The system combines a compact AI workstation with a large language model-based platform that converts enterprise documents into an interactive knowledge base.
The solution targets organisations that want to deploy AI while retaining control over internal data. By processing information on-premises, the platform can reduce dependence on cloud AI services and support data sovereignty, security, and regulatory requirements.
AiKMSpark uses Agentic AI capabilities to consolidate information from multiple enterprise documents and make it accessible through natural-language interaction. Employees can retrieve information, analyse data, generate insights, create structured reports, and export results in PDF or Word formats.
The AI BOX-A395 provides the local computing capability for these workloads. It is powered by the AMD Ryzen AI Max+ 395 processor, which integrates CPU, GPU, and NPU resources and delivers up to 126 TOPS of AI performance. The system supports up to 128GB of unified memory and can run large language models with up to 120 billion parameters locally.
The combination is intended for applications such as technical support, product documentation, regulatory compliance, and financial analysis. Local inference can also help organisations maintain control over AI models, enterprise information, and system operations while reducing infrastructure dependencies associated with cloud-based deployment.
The compact workstation is positioned as a ready-to-use platform for enterprise AI deployment, combining edge computing with document-based knowledge management. The approach is designed to simplify deployment for organisations moving AI workloads closer to where enterprise data is generated and stored.
Click here for the official announcement.
