For all the advances in technology that are impacting supply-chain management today, humans are still essential to understanding and fulfilling organizational goals, says Razat Gaurav, chief executive officer of LLamasoft.

SCB: How can companies can remain resilient and digitally transform their supply chains in the face of big innovations such as artificial intelligence, machine learning, and additive manufacturing?

Gaurav: Mathematical techniques and enabling technologies are providing computational horsepower, and creating a whole new level of the art of the possible. But when it comes to deciding the business model you're trying to achieve, the key pivots and transformations you need to make, it still goes back to the fundamentals of the business. If I'm a supply-chain organization, managing sourcing, production, distribution, and fulfillment, those functions haven't changed. Despite all the new technologies that are available, you’re still focusing on customer-service levels, on-time delivery performance, minimization of working capital, and balancing all of those across different constituents. With that in mind, enabling technologies can create a vision of the future and a roadmap that you're able to execute on.

There are some really interesting use cases we're seeing across the supply chain, whether in demand modeling, predictive analytics or inventory and transportation management. AI and machine learning offer another mathematical technique that we’re able to leverage. You still need to know what the business problems are that you're trying to solve, then map what the technology offers you, as opposed to doing it the other way around.

SCB: We become dazzled by science and technology, and forget that there are still human beings in the mix who have to make those decisions.

Gaurav: Right. One of the things that all this technology is still not able to provide is human judgment. We can make predictions and reduce costs through computational horsepower and algorithms. But there’s no substitute for people.

At the end of the day, you still need to collaborate. That’s more than just a mechanical exchange of data and information. As a manufacturer or retailer, you need to have relationships with your suppliers and carriers. Or if you’re a transportation organization, you have to work with warehousing, the finance team and marketing. The foundation of trust is still really important in making supply-chain decisions. Machines cannot replace that.

SCB: I thought that the original idea behind AI was to model and reproduce human judgment in a machine environment. It sounds like we haven't reached that point, and maybe won’t anytime soon.

Gaurav: The initial applications of AI and machine learning have all been focused around predictions. You might have three or four options, but deciding which one to pursue sometimes still requires human judgment. Often you’re having to made tradeoffs: Are you going to put more weight on quality? On cost or price? It's not a black box where you just plug and play. You need human beings with different skill sets to help you make those decisions.

SCB: What kind of skills and talent and do you need to carry out these tasks?

Gaurav: There’s a skill gap in the supply-chain profession of people who can understand the business context as well as the details and constraints of operational processes. And those who can tie that to the digital world of data and mathematical models. We need people who have both business understanding and a grounding in data science and mathematics. We're beginning to work with universities to influence their curriculums and programs for building the supply-chain professional of the future. Today’s academic programs aren’t really training students in how to bring these two worlds together, and converge them in a cross-functional, cross-departmental way.

SCB: Are universities open to your input?

Gaurav: They really are. We’ve found that they’re remarkably open-minded about understanding the challenges of the field. We’re working closely with the University of Michigan, and have started working with MIT, Georgia Tech and a few others. But even the universities are organized by different departments, each with its own sphere of focus.

SCB: So you have silos in education, just like in the business world.

Gaurav: That's right. We're trying to really bridge those silos and create cross-departmental curriculums that can get the operations research and industrial engineering departments to work with the business school, as well as the data and computer science departments.