Research work · 2026

From Data to Decisions Transforming Analytics with Artificial Intelligence

Data Analytics and Artificial Intelligence

Abstract

Artificial intelligence has moved from a peripheral tool in the analytics stack to the mechanism that now defines how organisations sense, interpret, and act upon data. This chapter traces that shift and argue that the conventional description of analytics as a four-stage progression - descriptive, diagnostic, predictive, and prescriptive - no longer captures what contemporary AI-enabled systems actually do. Rather than analysts querying data and interpreting output, increasingly the system itself proposes the interpretation, and in agentic configurations, initiates the resulting action. The chapter develops a framework for understanding this transition, examines the specific technical contributions of machine learning, deep learning, and generative AI to the analytics pipeline, and considers what is lost as well as gained when interpretive authority is delegated to a model. Specific attention is given to governance and explainability, not as norms afterthoughts but as structural requirements once decisions are automated at scale. Here are Five industry contexts - healthcare, finance, education, manufacturing, and urban infrastructure - are used to illustrate how these technologies behave differently once situated in a real operational environment, where data quality, regulatory exposure, and human trust vary considerably. The chapter closes by identifying open problems that remain underexamined in the literature: the evaluation of reasoning quality in generative analytics systems, the sustainability cost of large-scale model training for routine analytical tasks, and the question of accountability when an autonomous analytics agent acts on an inference that later proves wrong.

DOI: 10.46632/daai/6/1/3