Areas of Focus
By Sebastian Pokutta
Big data will finally reach maturity when you don’t see it. It will be everywhere in our lives, but it’ll be invisible. That’s where data is heading.
Right now, we’re at a stage where the information is there, and we’re trying to figure out what to do with it. People believe data is money, so most companies are harvesting data in very aggressive amounts. But so many companies still don’t know what to do with it. Google, Facebook, Microsoft and Instagram – those are the places where the business model is based on data. But, not many others have figured out how to work with it. The data is stored in different places, it’s not actionable, or it’s in an unstructured heap.
So, companies can get paralyzed by the huge amounts of data. And if there's a company not already working effectively with big data, it’s hard to hire top talent, because the top talent wants to go to Google and Facebook. Few are interested in working on machine learning and big data at a traditional, brick-and-mortar business.
And there’s a more fundamental thing: The companies that want to get into the game should hire a head of innovation who hasn’t been trained in, and doesn’t conform to, conventional standards. The worst thing these companies can do is miss the opportunity to hire a risk-taker as head of innovation simply because he or she would not conform to the typical board- and chairman-approved codes. You have to aggressively hire top, young talent for leadership positions to get ahead and to make innovation a key component in the corporate DNA.
For example, Macy’s understands that big data innovation is critical for success. They’ve sponsored various big data projects at Georgia Tech, and we designed solutions for their needs. One solution was prepackaged inventory, or using data to predict what people will be ordering, so that they can pre-package the items and then slap an address label on them.
This helps balance worker demand. At peak times, they’re sending, and in slow times, they’re packaging.
When you have erratic demand, you want to avoid having your workers busy one day, but not the next. You want to reassign workload.
Where’s it all going? The only reason businesses collect all this information is to make decisions. In the future, you’ll see more autonomous systems as you go through your day. They’ll be processing very large amounts of data immediately to make decisions for us.
Sebastian Pokutta works at the intersection of theoretical frameworks and practical applications of big data. As the David M. McKenney Family Associate Professor in the Stewart School of Industrial and Systems Engineering (and an associate director of Georgia Tech’s Center of Machine Learning), he has led more than two dozen research projects in big data, machine learning, and optimization with a wide range of companies.