Allganize Launches MCP-based 'Alli Agent Builder'

## Allganize Unveils Security-Enhanced 'Agent Builder,' Opening New Horizons for Building Customized Enterprise AI Agents Allganize, an LLM all-in-one solution company, has elevated the safety and efficiency of enterprise AI utilization to a new level by integrating the innovative 'Agent Builder' feature into its 'Alli' platform. This is regarded as a core solution that resolves the challenges of building MCP (Model Context Protocol)-based AI agents, which demand high-level security as much as powerful performance. MCP is a cutting-edge technology essential for LLMs to flexibly call various functions and data to automate complex tasks. However, when externally developed MCP components are executed directly by AI, they demand extensive permissions, leading to a structural characteristic where the stability and security threats to enterprise systems are inevitable. For this reason, despite execution in a completely isolated and controlled environment being essential, the common industry consensus has been that systems satisfying practical security requirements are rare. Based on its accumulated experience, having developed enterprise AI since 2017 and applied generative AI to enterprise settings since 2022, Allganize has perfectly resolved these security and operational challenges. The 'Agent Builder' is implemented to safely execute MCP-based structures within a completely controlled environment, allowing enterprises to maximize the potential of generative AI agents while alleviating security concerns. This solution guarantees equally excellent performance in both SaaS and on-premise environments. Especially in on-premise environments, it maximizes performance by utilizing Allganize's self-developed optimized LLM models, 'Alpha-V2' and 'Alpha-R1,' and operates perfectly even in infrastructures with stringent security policies applied, such as external network blocking and data access restrictions. Furthermore, agents can be easily configured without coding, and by linking with the 'Agentic RAG (Retrieval-Augmented Generation)' feature, it is possible to perform in-depth reasoning from complex documents and data. AI agents created with Agent Builder are expected to be applied across various enterprise work settings. Examples include 'Search Agents' that perform integrated searches of internal and external data, 'Data Analysis Agents' that analyze complex BI (Business Intelligence) systems and automatically generate reports, and 'Sales Support Agents' that analyze the latest news and documents to summarize key information and formulate sales strategies for meetings. Allganize currently holds approximately 300 enterprise clients primarily in Korea, the US, and Japan, across sectors such as finance, public services, and manufacturing. Through the Alli platform, it provides all the necessary functions for enterprises to utilize generative AI, from industry-specific on-premise AI models to Agent Builder and Agent Market, thereby leading the market. Allganize CEO Lee Chang-soo stated, "Many companies are embarking on building MCP-based agents, but systems that fully satisfy actual security requirements are rare." He emphasized, "Based on our extensive experience and accumulated technological expertise, Allganize will provide complete AI agents that meet the high-level security and operational requirements desired by enterprises." This is expected to pave the way for enterprises to confidently adopt the innovative power of AI agents into their operations.