3 Ways CIOs Can Shepherd Machine Learning and AI Adoption in the Enterprise

PinIt

A CIO can often be cast today in an advisory role, helping the entire enterprise, rather than a single department, to improve KPIs.

The role of chief information officer is changing, as organizations shift their digital transformation strategies from a focus on data to a focus on machine learning and artificial intelligence.

No longer can CIOs decide *if* they should use AI, but instead where and how to apply it. As a result, CIOs are facing increased pressure to take on a more advisory role, helping the entire enterprise, rather than a single department, improve their KPIs by embracing AI-driven technology. 

See also: How CIOs can act as a growth driver

The application for AI-driven decisioning spans across departments, including marketing, finance and human resources. The CIO should be guiding adoption and execution throughout the organization in myriad use cases to drive better outcomes across the business landscape. It’s not enough to just collect information, then analyze noise in addition to valuable data, and make rear-view mirror predictions about what to do next. The CIO should be helping the organization to make better use of their data and continually improve decisioning based on AI learning from past outcomes.

Here’s how to ensure AI adoption is successful across the business:

#1: Get AI into the Hands of Business Stakeholders

CIOs should ensure that the powerful tools being used are accessible by more than just a few analytic gurus in the organization. If that is the case, it’s more than likely that the usage will be confined to a few initiatives and won’t scale to the meet the needs of the business across departments. Perhaps more importantly, such a process will leave out the knowledge of domain experts. This knowledge is critical to the success of any initiative and should not be overlooked. In fact, the programs with the greatest success put the AI tools directly into the hands of the business stakeholders.

#2: Ruthless Prioritization: Program Management not Science Experiments 

If an organization is trying to incorporate AI as part of the digital transformation journey, the primary goal is to find a faster, smarter way to get from data to action. This isn’t going to happen overnight and will require prioritization and focus to be successful. The CIO can bring a wealth of experience with iterative processes – the foundation of agile development methodology – and can help the organization prioritize AI implementation, focus on generating success in those areas, and bring consistency across departments. 

Using a diagnostic medicine supplier as an example, an organization may want to implement analytics, AI and machine learning to manage its demand planning process. In this case, the real challenge was identifying ways to adapt technology to the diagnostic medicine business and integrate the analysts who manage demand daily with deep business knowledge. Once that solution was successfully implemented, the supplier reduced the time to create a more accurate sales forecast from one month to 30 minutes. It also prompted the supplier to consider ways to expand AI into other applications. The land-and-expand approach allows each succeeding application to occur faster — with fewer issues, more collaboration across silos, and ultimately faster return on investment. 

#3: Understand the Why: Make AI Explainable

The days of “black box” AI are numbered. As AI systems are given more responsibilities and more complex tasks, business leaders must understand the basis of the decisions so that they can be trusted.  If a business is making important bets on a machine’s decision, it’s critical to understand why the machine is making a certain decision and on what basis the decision is being made. Increasingly, the CIO will need to work with business leaders to describe the decision-making process, understand the actions taken as a result, then measure the effectiveness of those decisions and actions.

Explainable AI should produce more transparent models and maintain a high level of predictive accuracy, while enabling users to understand, trust and manage the system. This is particularly important when AI-derived decisions impact consumers.

One clear example of its importance is found in Article 22 of the General Data Protection Regulation (GDPR). It grants consumers the right “not be subject to a decision based solely on automated processing.” It further provides consumers the right to obtain additional information about a decision reached on the basis of automated processing and to contest this decision. Explainable AI can provide the insight on how the decision was reached and help ensure that the process was not discriminatory.

Nearly every business is figuring out how to use AI and machine learning to better compete. This means CIOs are now responsible for leading the organization through the journey and shepherding the employees successfully through the process. Starting with a clear grasp of where the business stands now – employees, customers, assets and infrastructure, CIOs should understand the vision of where the business wants to be and create the roadmap to get there. While the journey will have small steps and huge leaps, dotted with surprising discoveries and a few failures, the CIO can lead the organization with the right blend of people, data, analytics and machines to a future that is uniquely its own.

Bill Waid

About Bill Waid

Bill Waid is chief product and technology officer for FICO. In this role, he drives development of the FICO Platform, which provides the ideal decisioning foundation companies need to successfully achieve digital transformation. Bill previously was general manager for Decision Management Systems at FICO, helping firms adopt analytics, decisioning, and optimization solutions that identify and take action on unique, predictive insights in real-time. Bill joined FICO in 2002 to lead the formation of the decision management business, now one of FICO's largest and fast-growing offerings, serving thousands of customers in a wide range of industries across the globe. Bill began his career as an engineer for Neuron Data and Blaze Software. He holds a civil engineering degree from Lehigh University.

Leave a Reply

Your email address will not be published. Required fields are marked *