And at the moment, those experts come from academia. There’s a misguided assumption that AI is data science. It’s not (although data is the fuel). To truly understand the complexities of these algorithmic models and to customize them to your business needs, you need someone trained at the forefront of machine learning and AI. Every AI company – from consultancy to large enterprise – should have at least one PhD-level researcher on their team, whether they’re directing research or scoping technical architecture. No matter how smart your IT team is, the field is too new and the expert talent pool too small to leave the matter entirely in inexperienced hands.

“In the sales process, we often see that there’s a person in the company who really wants to get into machine learning and is pursuing it as a hobby or side project,” says Dr. Oates. “And typically, that doesn’t end well. The person has to know a lot about the domain and have the depth of knowledge to come up with really good solutions to hard problems.”

A good AI consultancy will also have strong business leadership with a proven track record. Successful outcomes will be driven by the marriage between academia and industry. “I’m good if someone brings me an input and an output and wants to know how to get in between the two,” Dr. Oates continues, “But you also need someone who can figure out the right way to deliver value that the company recognizes and that’s Stephen. That’s the complete package.”