With multiple layers of suppliers, artificial intelligence and big data can help companies get clean and useful information, says Jim Hayden, chief data scientist at Everstream Analytics.

AI and big data are solving issues across the supply chain, from planning to manufacturing to transportation, Hayden says. “It's emerging, and companies are competing on it today. If you're not applying machine learning to your data, you're falling behind.”
 It isn’t easy to get to the truth, he notes. “It takes different data sources to zero in on it. As an example, you need to identify your suppliers and their suppliers. To do that, you need to know their name and their location. There are many different sources that can tell you that. There are some high-quality data sources that you can use for reference. And then there's the operational data. That's often a little dirtier, but when you combine the two, you can get to the truth.”
Understanding your sub-tier suppliers is a major challenge in supply chain. “You do business with your tier-one suppliers, but you're not sure who they do business with — and who they do business with. 

The key lies in the ability to capture import-export data. “That's a record of who's trading with whom,” Hayden says. “We have billions of those records that we use to then derive those sub-tier relationships. That's not a very clean data source, as you can imagine, coming from hundreds of different customs systems from around the world. And that's an example where we need machine learning technology to truly identify who's involved in these trades. It's called entity resolution, and resolving the entities involved is a tricky problem that requires pretty advanced technology.”

Bad data means bad decisions. And that can equate to loss of money and reputation, Hayden says.