A Unified Framework for Agentic AI and Data Products: Enhancing Cloud, Big Data, and Machine Learning in Supply Chain, Insurance, Retail, and Manufacturing

Abstract

While the evolution of deployed AI technologies is attracting increasing interest and investment across domains, there remains severe ambiguity on the question of what types of technologies should be categorized as “AI”. Indeed, there exists an entire class of digital technologies in production that utilize a variety of different algorithms, ROI-base value propositions, and user experiences, some developed with AI techniques and some leveraging statistically-defined heuristics and decision support. This paper proposes a framework to help clarify the boundaries between algorithmic, AI, and agentic AI systems as well as how these systems may be utilized in pursuit of different data product goals. We believe these discussions can lay the groundwork to support industry-specific discussions on this important topic going forward.