Like it or not, AI will soon infiltrate every aspect of our society from our refrigerators to autonomous cars and everything in between. Combined with data and data science, AI offers new, smart ways to solve existing problems and create fresh and exciting opportunities. But how do you determine the importance of an AI initiative for your organization and how do you ensure you derive the maximum benefit from implementing such projects? By identifying business needs and viewing AI and data science as a means to fulfilling those needs, IT professionals and data scientists can communicate the value and transformational affect AI can bring to the organization.
To begin, think of AI as an experiment rather than a pilot project. Unlike other technologies, every application of AI requires different tools and algorithms. Experimenting with different AI applications is typically done repeatedly as you explore the many options. Experimenting enables you to understand whether you can solve a problem more effectively, with better, more predictable outcomes, and transform a business process or function using AI.
An experiment may not proceed for any number of reasons, but that’s not an indication of failure. Learning what works and what doesn’t by experimenting can accelerate your AI journey. For example, one pharmaceutical company experimented with natural language processing and voice analytics in its patient services call center. Automating a process to alert managers to patient frustration enabled the managers to take the appropriate intervening steps to keep patients on critical drug regimens. While the experiment was a success, the technology required a huge transformation of the patient services manager workforce, including up-skilling or replacing some of the staff. This created a significant barrier and prevented the company from taking the pilot project into production.
On the other hand, data is what fuels AI; however, the sheer quantity of data organizations must capture and manage is overwhelming, and every organization will need to determine which data is most important. How do you separate the data you want from the data you don’t? Before embarking on an AI solution, identify what is driving the business need or opportunity. Apply AI, machine learning or design thinking to help understand those drivers and identify core and missing data. For example, applying machine learning and design thinking to existing sales data will reveal what drives target key performance indicators (KPIs), how that increases conversion rates, what directly correlates to those KPIs and what core data is most important. Machine learning will enable you to solve the problem in a smarter and more effective way by embedding algorithms and models that can automate and transform the entire business process. This exercise will help the organization approach the change needed to adopt AI more holistically.
Unique in the world of tech evolution, AI offers unprecedented capabilities. AI will deeply impact the future of work, affecting employees, customers, internal processes and more. Understanding these impacts and preparing for change is even more important than the power of the technology itself. It’s critical to create an environment that allows for autonomy and, simultaneously, experimentation. This needs to be balanced by ensuring AI is a shared capability to prevent reinventing the wheel each time a project is proposed. Bringing together knowledge from the business, a function or process, data and AI technology, and developing a governance structure where everyone works in unison, will help foster successful AI initiatives. Approaching AI as a shared capability will enable you to develop an ecosystem of partnerships, generate common best practices and create reusability. While organizations that work in silos will find themselves spending considerable time, effort and money playing catch up.