Artificial Intelligence Operating Model: A Proposal Framework for AI Operationalization and Deployment

Abstract

At the heart of the newenterprise, across all activities, is a decision factory governed by some kindof intelligence. Among the great promises of Artificial Intelligence (AI) isits ability to lead to a significant evolution in the amount of data received,processed, or generates by companies, particularly those with a digitalconnotation. To bring about dramatic changes, AI does not need to be sciencefiction but simply a new way of approaching computerization subjects whether interms of design, development, or terms of expected results. It should be notedthat traditional IT solutions present a form of AI called - Weak AI - while theAI that is the subject of much noise, hype, and promises of transformation andpotential for growth is called - Strong AI -. This article aims to present, ina didactic way, a model called D2MO (For Data Ops, ML Ops, Model Ops, and AIOps) allowing the company to operationalize, in a structured approach, AIsubjects, activities, and projects. We target through this article to provideboth IT and business experts with a new framework offering a perfectarticulation between the different bricks and actors entering into thecomposition of an AI-based system thus allowing them to operate in harmony andan agile mode while taking advantage of this technology.