AI-Native Cloud Platforms: Redefining Scalability and Flexibility in Artificial Intelligence Workflows

Abstract

Cloud computing and AI have conventionally been delivered as distinct services. Over the past few years, however, they are increasingly becoming integrated and integral components of a wider service offering. This essay focuses on the intersection of AI and cloud and introduces AI-native cloud platforms—platforms that are developed with AI, big data, advanced analytics, and machine learning as a ‘default’ integrated part of the platform design.

Founded on the underpinnings of AI as a service and cloud-native AI, this essay argues that AI-native cloud platforms represent the ‘future’ cloud platforms. We introduce scalability and flexibility as cornerstones of AI-native platforms and show how these platforms deal with key challenges plaguing the current AI platform frameworks. We draw from a case study to showcase real-world cloud-based AI-native platforms. The case example in this essay provides a good theoretical and practical lens for understanding the cloud as AI and portends a future world where AI becomes integrated into, rather than encapsulated from the cloud. Thus, the propositions we present will be of interest to both IS researchers who are interested in cloud and AI as well as to practitioners who may be involved in the design and development of contemporary data platforms and software applications.

With opportunities in the environment surrounding AI-native cloud platforms, this essay seeks to move beyond both the developments and the challenges unique to the cloud as AI. At the same time, we argue that looking at AI as part of cloud computing will enable us to begin addressing some of the contemporary issues in AI that are associated with development in the ‘other’ direction, i.e., from data and AI models out to real-world applications.