Four Quadrants of the Enterprise AI business case

In this post, I discuss the development of the Enterprise AI business case through a framework of four quadrants.  According to Gartner: “The mindset shift required for AI can lead to “cultural anxiety” because it calls for a deep change in behaviors and ways of thinking”. Deployment of AI in an Enterprise is complex and multi-disciplinary. Hence, this framework is evolutionary.  The vendors and initiatives listed are included to illustrate the framework.

In progressive orders of complexity (and opportunity) the four quadrants for the Enterprise AI business case are:

The analysis is based on the Enterprise AI workshop in London and remotely

Enterprise AI is an abstract concept, interdisciplinary and much-hyped concept. But in any case, the deployment of AI in the Enterprise cannot be viewed in isolation. Within the Enterprise, there already exist systems (like ERP and Data Warehousing). The integration of these will have a role to play in any AI deployment. The word ‘Enterprise’ can be seen in terms of Enterprise workflows. We also consider the core Enterprise (a non-manufacturing company ex Insurance) and the Wider enterprise (including supply chain). Hence, Enterprise AI could be understanding how workflows change when AI is deployed in the Enterprise. 

The professional deployment of AI in Enterprises differs from the content in a typical training course. In larger organisations, the Data Science function typically spans three distinct roles: The Data Engineer, the Data Scientist and the DevOps Engineer. The Data Scientist is primarily responsible for developing the Machine Learning and Deep Learning algorithms. The Data Engineer and The DevOps Engineer roles work in conjunction with the Data Scientist to manage the product/service lifecycle. Hence, in an Enterprise, managing the AI pipeline involves the philosophy of CICD (Continuous Improvement – Continuous Delivery). CI/CD can be seen as an evolution of Waterfall and Agile methodologies.

Finally, to clarify some definitions used: Machine Learning: Means systems that can learn from experience (Data); Deep Learning: Implies a system that can perform automatic feature detection based on Deep neural networks; Artificial Intelligence involves machines that can reason.

With this background, let us explore the four quadrants of the AI business case

We could initially model the problem as a machine learning or a deep learning problem. At this stage, we are concerned with the accuracy, choice and the efficiency of the model.  Hence, the first quadrant is characterized by experimental analysis to prove value.

We are also concerned with improving the existing KPIs. For example, if you are working with fraud detection or loan prediction – each of these applications has an existing KPI based on current techniques. The machine learning and deep learning models would be expected to significantly improve the current benchmarks. We are typically working with one node(non-distributed) processing. The Data could be in Time series, Tabular, Textual, Image, Audio or Video based. The applications could involve Computer vision, NLP, Fintech/financial services, Healthcare, Reinforcement learning, Unsupervised learning (ex GANs, VAE), Emotion AI (Affective computing) etc. Deep learning architectures are rapidly evolving.  Hence, there is a lot of effort and skill needed at this stage.

Building on from the first quadrant, the second quadrant is characterized by

Both ERP and Data Warehousing exist in large Enterprises.