AI Readiness & Performance: Doing Things Right & Doing the Right Things

“Managers do things right. Leaders do the right thing.”   

That observation is credited to author and USC business professor Warren G. Bennis, and it succinctly reminds us that we need both skills and vision, execution and strategy, in whatever business we undertake.   And it posits that some of us are better at one or the other.

As public and private sector organizations worldwide race towards digital transformation and the productivity boosts and competitive advantage that can derive from data science and AI, it is important for executives and managers to ensure that they’re doing the right things AND doing them right.

AI Readiness – Doing It Right

The core of successful, transformative Artificial Intelligence (AI) is machine learning (ML).  The essence of ML is the mastery of three key stages: 

• Acquiring, storing, protecting and preparing data
• Analyzing that data and building models that describe, classify or predict 
• Deploying those models and integrating the models’ outputs into business, operations or  engineering actions that achieve an organization’s goals. 

An organization’s effectiveness in any undertaking is affected by the level and quality of inputs such as policies and processes set by management, available infrastructure, and the skills of its people at different levels and in various departments and roles.

"We are what we repeatedly do. Excellence, then, is not an act, but a habit." - Will Durant

Thus true AI readiness consists in – and AI performance depends on – robust skills and processes for everything from data acquisition, cleaning and warehousing through model validation to the stage-gating of solutions to be developed and deployed. These capabilities can be assessed and benchmarked against industry standards and against an organization's earlier level of readiness. Identified gaps can be addressed through training and through wise upgrades to hardware and software infrastructure.

Project Identification and Selection – Doing the Right Things

Suppose you’ve mastered those skills and inculcated those best practices.   Great – your organization has the tools to do competent AI and data science! But where and when should you deploy it?  For which applications, which strategic objectives?  To move the needle on which of your organization’s KPIs?

There are several ways to approach this.   I like to use an iterative, consultative process of ideation and strategic filtering to identify and rank potential projects.  Ranking and filtering criteria include the organization’s strategic goals – e.g., gaining market share with a new product feature, or lowering operating costs through warehouse automation – as well as data and performance requirements, budget, projected ROI and ethics considerations (bias, privacy, transparency, safety). Ideally this process brings in people from different functions across the organization, and MUST include a customer or end-user!

"Empowering innovations transform something that is complicated and expensive into something that is so much more simple and affordable that a much larger population can enjoy it." - Clayton M. Christensen

All components of organizational readiness need to be strong for sustained success in AI strategy and implementation.  For example, a great application idea and powerful machine learning algorithms are of little use without the necessary data for training and validating the predictive model.  And none of it matters if customers or employees can’t use the deployed system with confidence and trust -- so UI/UX, rollout, communications, documentation and training matter.

I’m fortunate to be working with a group of data scientists, engineers and successful serial entrepreneurs who excel at all phases of AI development and deployment, and who provide AI readiness assessments, customized training and solutions to organizations throughout the MENA region.

Evan Steeg is an innovation consultant in the Toronto-Ottawa-Montréal region, and is Head of Client Solutions at Ottawa- and Dubai-based Stallion.AI. He has a Mathematics degree from Cornell University and a PhD in Computer Science from the University of Toronto.