State owned Assets Recognition! SIE Information was invited to participate in the "Modern New State Owned Enterprises" seminar at the State owned Assets Supervision and Administration
On April 24th, the "Modern New State Owned Enterprises" seminar was held in Hengqin, Guangdong, focusing on the theme of "Application, Data, and Intelligent Computing: Connecting the Last Mile of Enterprise AI+" to explore the empowerment path of domestic large-scale models. This seminar is jointly guided by the Executive Committee of Hengqin Guangdong Macao Deep Cooperation Zone, the News Center of the State owned Assets Supervision and Administration Commission of the State Council, and the State owned Assets Supervision and Administration Commission of Guangdong Province. It is hosted by the "State owned Assets Report" magazine, the Economic Development Bureau of Hengqin Guangdong Macao Deep Cooperation Zone, and China United Network Communications Group Co., Ltd. It is organized by the Zhuhai Investment Service Bureau, Hengqin Shenhe Investment Co., Ltd., and Guangdong Zhiyong Artificial Intelligence Application Research Institute. More than 400 experts and scholars in the field of artificial intelligence, representatives of state-owned enterprises and well-known private enterprises, etc. participated in the forum.
SIE Information, as the only private AI product enterprise invited to speak at the manufacturing industry session, deeply analyzed the implementation path of enterprise level AI applications and officially released the new generation of full stack enterprise level AI middleware - "Shanmou GPT 2025". Together with the hosting unit Intelligent Artificial Intelligence Application Research Institute, SIE Information provides enterprises with more stable, efficient, and low threshold AI integration and scenario based application solutions, accelerating the large-scale implementation of intelligent transformation scenarios.
Liu Ziheng, Vice General Manager of SIE Information and Secretary of the Board of Directors, delivered a keynote speech at the seminar, systematically elaborating on the underlying logic of enterprise level AI application implementation, the similarities and differences between decision based AI and generative AI in enterprise management and manufacturing fields, revealing the difficulties and success conditions of AI application implementation, and ultimately providing a set of hierarchical guidelines for enterprise level AI application implementation solutions. SIE Information firmly believes that the solution to the implementation of enterprise AI applications is to build AI application intelligent agents through the AI toolchain platform, which provides new ideas and methodological references for the practical exploration of enterprise level AI technology in the state-owned asset system.
In response to the three core pain points of data fragmentation, high compliance costs, and slow scene implementation faced by state-owned enterprise transformation, SIE Information has built a full stack AI solution consisting of a "technology platform+scene ecology+engineering capabilities":
Technical middle layer: Based on the self-developed AI middle platform Shanmou GPT, it integrates core capabilities such as MCP protocol (model context protocol), A2A technology (automation proxy collaboration protocol), multimodal data governance, and knowledge graph construction to help enterprises build AI application intelligent agents with building blocks.
Scenario Ecological Layer: Deeply embed intelligent agent technology into core business scenarios such as marketing, manufacturing, supply chain, finance and taxation, and human resources, achieving full chain intelligence from data organization to decision support.
Engineering Capability Layer: Provide engineering services such as low code development, localization adaptation (supporting private deployment and innovation environment), agile iteration, etc., to ensure the rapid implementation of AI applications.
Shanmou GPT 2025 is newly launched
Assist state-owned enterprises in building AI capabilities
As the new generation AI platform for SIE Information, "Shanmou GPT 2025" innovatively integrates MCP and A2A core technologies, breaking through the integration barriers of traditional AI applications and providing a powerful engine for the intelligent transformation of state-owned enterprises
MCP protocol connects the entire system: Deeply supports MCP protocol, seamlessly integrates mainstream models at home and abroad such as DeepSeek, Llama, Tongyi Qianwen, etc., to achieve full chain collaborative optimization and efficient operation.
Visual Agent Construction: Based on A2A technology, pre-set agent templates are used in fields such as marketing, finance and taxation, and manufacturing, covering scenarios such as data organization, report generation, and customer service. Workflow automation can be achieved without coding.
Efficient knowledge utilization: Supports structured management of documents/images/videos, combined with RAG technology to generate knowledge graphs, accurately extract information, and convert it into business insights.
Flexible workflow orchestration: Non technical personnel use zero code tools to build multimodal intelligent agents for scenarios such as data organization, to meet complex business needs and improve efficiency in task processing.
SIE Information's new generation AI platform "Shanmou GPT 2025" strengthens AI collaboration capabilities with MCP protocol and uses A2A technology to assist state-owned enterprises in efficiently building exclusive AI applications. Through a "fast, simple, and low-cost" lightweight deployment model, it effectively solves the problems of complex integration, high development threshold, and slow scene implementation in intelligent transformation, and activates the potential for business innovation in state-owned enterprises.
In the future, SIE Information will continue to deepen the integration and innovation of AI technology with industry scenarios, provide intelligent solutions that meet the business needs of more state-owned enterprises, continue to promote the construction of more intelligent agent applications, comprehensively empower enterprises to improve quality and efficiency through "enterprise application intelligence", and build a better new future for digital enterprises.
