ChinaSoft International (00354) has announced a key advancement in its allmeta product suite with the official launch of allmeta Ontology v1.0. This platform is designed as an actionable ontology system to facilitate the practical implementation of enterprise AI in real-world business scenarios.
The newly released allmeta Ontology v1.0 addresses the common challenge where large language models fail to grasp a company's deep-seated business rules. It features six core builders—Object, Action, Rule, Event, Relationship, and Evaluation—that automatically transform data from systems like ERP and MES, along with unstructured documents, into actionable knowledge blueprints for AI agents. This process enables AI-controlled execution and full-process code generation without the need for manual workflow orchestration or prompt engineering.
A central innovation of the platform is its AI-driven approach to both ontology generation and runtime operation. It manages AI agents as code, significantly reducing the engineering costs associated with migrating from traditional systems to an AI-native architecture. Additionally, the built-in Agentic Operator serves as a runtime management and orchestration platform, supporting event-driven dynamic scheduling, unified governance, and permission controls. This ensures that production-level AI is traceable, controllable, and allows for immediate logic corrections.
In collaboration with ChinaSoft International's Field Delivery Experts (FDE), allmeta Ontology v1.0 has already been deployed in various scenarios, including power grid dispatching, recruitment automation, and expense reimbursement. The platform is compatible with leading large language models globally, assisting enterprises in genuinely achieving "using AI to build AI" and entering a new phase of "Agent as a Service."
The company views allmeta as a crucial piece of infrastructure for the enterprise AI era. The release of allmeta Ontology v1.0 marks a significant step forward in evolving corporate AI from simple question-answering to actionable execution, from intelligent dialogue to intelligent operations, and from model capability to business operational capability. This progress is expected to lay a solid foundation for the group's future exploration and transformation towards business models centered on high-value professional token operations.
Comments