Six Quick Tips for AI Project Success

Tony Brownlee
10/4/16 6:30 AM

We've seen continued focus on cognitive, or artificial intelligence (AI) technologies from some of the larger tech companies in the world. This Machine Learning Tipswave of technology is truly revolutionary, but it can be a bit confusing. With new definitions popping up like a frenetic game of buzzword bingo, I prefer this simple explanation used by New York Times reporter Quentin Hardy: 

"Cloaked inside terms like deep learning and machine intelligence, AI is essentially a series of advanced statistics-based exercises that review the past to indicate the likely future."

As I've said before, I think everyone should be taking on an AI project. As you're working on your plans, here are six tips for understanding this space and narrowing your priorities.

  1. Large and smaller vendors are building strong AI capabilities. While the New York Times article references firms like Salesforce, IBM, and Oracle, there are many smaller technology companies that are building expertise with machine learning, AI, and related technologies. At Kingland, these AI capabilities are the number one area of investment in our Kingland Platform, as a critical series of components focused on data management and compliance processes.  

  2. Look for the data. From banks and insurance companies, to energy companies and retailers, data quality is important. As data volumes continue to grow, exception volumes grow as well, which leads to growing processing costs. Data processing use cases make great AI projects. Millions of records, hundreds of attributes, thousands of rules... and typically a lot of annual expense.   

  3. AI is not one-size fits all. Think of it as a doctor. You wouldn't hire a brain surgeon to operate on your heart, and you wouldn't go to a cardiac specialist for a broken bone. Look for partners who are specialists in a particular business domain, industry, or data type AND have a committed focus on artificial intelligence.  

  4. AI isn't magic, it requires the right skills. I've spent the last number of days with thousands of college students at one of the largest engineering career fairs in the United States. Resoundingly, the top software engineering talent that we're talking to 1) loves the idea of working with AI technology and 2) is coming out of college with skills that make them productive day one. Skills can come in many shapes and sizes, but it's important to have an explicit focus on attracting, developing, and retaining the individuals with these capabilities.  

  5. Set realistic, yet big expectations. There are two ways to think about the benefits in your mind - progress, and game-changing. Some AI projects are going to be all about efficiency - taking tasks performed by 100 people and allow those same tasks to be performed by 50 people instead. Other AI projects set the vision on game-changing improvements - the types of improvements or new capabilities that simply weren't possible before. You need both...but should have different expectations about the time-frames, resources, and impact of each.  Set your expectations high in both.  

  6. Focus on business impact. The excitement around AI is the potential for those game-changing innovations. However, there's risk that it's too "out there" and unpredictable to be viable for the challenges that are getting prioritized in annual goals and key initiatives. Progress beyond this by keeping your AI project focused on real business impact, and then promoting that impact. Showing your fellow leaders and executives the types of improvements, efficiencies, or insights helps build momentum in the "art of the possible."   

These six tips can help guide you toward focusing on data that reveals a likely future. But to start showing both artificial and practical intelligence, you'll want to involve the right people.

Learn how we automate data processes with artificial intelligence

You May Also Like

These Stories on Text Analytics

No Comments Yet

Let us know what you think