T/IQA 12-2021 Intelligent Microservice Adapters Framework
T/IQA 12-2021 Intelligent Microservice Adapters Framework
Basic Information
Scope
Scope: This document specifies the use of container technology to encapsulate artificial intelligence models into intelligent microservices, form intelligent microservice combinations, and then integrate them into a framework of standardized intelligent business interface definitions, deployment strategies, and process combinations. This document is applicable to the R&D, operation and maintenance, and use of intelligent microservice products and applications in the field of artificial intelligence. Main technical content: This document is applicable to the R&D, operation and maintenance, and use of intelligent service products and applications in the field of artificial intelligence. The main practical content includes: 1) Standardizing interface definitions: Encapsulating AI models and algorithms into intelligent microservices using container technology, and standardizing the interface definitions of each intelligent microservice component; 2) Standardizing deployment strategies: Standardizing the strategy configuration and corresponding optimization methods for the deployment of multiple intelligent microservice components in the process of combination deployment; 3) Standardizing process combinations: Integrating the combined microservices into a complete standardized intelligent business process, serving the modern service industry, and achieving better management and maintenance. The intelligent service adaptation framework should support the encapsulation of AI models from different AI model libraries with different compilation and run environments into intelligent microservices, and the integration and combination of heterogeneous intelligent service models in an open intelligent service environment to form intelligent service processes, while achieving continuous integration and continuous deployment (CICD). The microservice adaptation framework should support mainstream AI algorithms, mainstream machine learning models, and mainstream machine learning en