In From the Data and AI Expert Labs, I’ll take the expert learnings gained through years of advising organizations to get their AI projects up and running and serve them up for you in an actionable and insightful way.
IBM’s Data and AI Expert Labs help guide organizations on their AI journeys. Expert Labs consultants represent years of data and AI expertise and offer best practices and business strategies to meet each organization’s unique needs.
Start with Outcomes
When clients reach out to us for assistance in transforming their enterprise, they often think the first step is deciding what AI product they need to adopt. But really, the first step any company needs to consider is “What is the business outcome I want to achieve?”
Think about AI not as a thing that automatically turns chaos into order, but as a tool that allows your business to solve problems and achieve a desired outcome. If you don’t know what that outcome is at the outset, your project will fail.
At the Data and AI Expert Labs, we’ve developed a formula to contextualize our outcome-first philosophy:
Outcome = Technology + Skills + Methodology
Let’s break that down. What clients want out of AI is to generate a specific outcome for their organization, and they want to scale that outcome. With that in mind, it becomes vital to think about how the technology you use will drive those outcomes. Suppose you’re in a customer service industry. In that case, you’ll need a robust CRM system and a virtual assistant to enable your customers to self-serve and to allow your employees to find correct information faster. If your industry is healthcare, then perhaps you’ll be looking at AI to ingest vast amounts of data to aid professionals with ML models that accelerate disease prevention, prediction, or cost of care.
AI without people is useless. In addition to having access to the technology, you’ll have harvested the skills and expertise represented in your organization. AI skills and expertise are not isolated to the data scientists in the organization, but instead to every job function that touches the AI application — from the end users in the line of business, to the application developer, the production management, etc. Your team will need to know how they can assist and be assisted by AI.
In many cases, the AI solution will handle rote tasks, but it will require a helping hand to get to that point. AI is a quick study, but it will always need people to make sure it’s learning the right things, and your team’s skills are essential to seeing success in that area. AI can’t simulate skills that aren’t in-house already. If your organization lacks specific expertise, it’s vital to shore up those areas by hiring or contracting until all needed skills are accessible to your organization.
Last is methodology, the processes on which your organization relies. As I mentioned earlier, AI is excellent at taking on tedious routine tasks, so your workflows and business processes will be streamlined, but you won’t understand what the effect of that will be if you haven’t thoroughly considered your outcomes. Once you’ve decided on your desired results, you can then start building a smart data infrastructure to pave the way for AI. Because AI needs good data, you must thoughtfully approach how your information architecture is organized and ensure your outcomes will be supported.
It’s essential to think about technology as a driver for results and enable people to use AI to close skill gaps and refine methodologies. By employing this formula, clients gain insights into how they can reach outcomes more quickly; then they’ll start to understand how to scale AI throughout their organization.
In my next post, I’ll introduce you to the IBM Data and Expert Labs POV on AI as a tool for transformation. Come back to the IBM Watson blog next week to read on.
In the meantime, if you’re interested in seeing how AI is able to rapidly scale and provide useful information during the COVID-19 pandemic, register for this valuable webinar on how to turn challenge into opportunity, presented jointly by IBM Watson and analyst firm Frost & Sullivan.
