Azure OpenAI Service for Cloud Native Applications

Chapter 2. Designing Cloud Native Architectures for Generative AI

Cloud native architecture is a way of designing and building applications that can take advantage of the cloud’s unique capabilities and constraints. Cloud native applications are typically composed of microservices that run in containers, orchestrated by platforms like Kubernetes, and use DevOps and continuous integration and continuous deployment (CI/CD) practices to enable rapid delivery and scalability. Cloud native architectures are at the core of the generative AI era.

Organizations such as the Cloud Native Computing Foundation (CNCF) are great catalysts of cloud native best practices and community development. Their goal is to be “the vendor-neutral hub of cloud native computing, to make cloud native universal and sustainable.” CNCF is a great source of information and learning material for these topics. Another great resource is the twelve-factor app, a public methodology for building cloud native applications.

As part of the cloud native movement, there are several projects and communities oriented to the use of cloud native architecture to enable scalable, reliable, and robust AI systems. They often require large amounts of data, complex algorithms, and specialized hardware to perform tasks such as image recognition, natural language processing, or recommendation systems. This is not always possible with traditional IT architecture patterns (e.g., monolithic applications).

The need for cloud native architecture ...