LinkedIn is working to train all of its engineers on the basics of implementing artificial intelligence as part of the company’s drive to make its professional social network smarter.

“The demand for AI across the company has increased enormously,” Deepak Agarwal, head of artificial intelligence at LinkedIn, said during an onstage interview at VB Summit 2017 today. “Everyone wants to have AI as a component of their product.”

The company has launched an AI academy for its engineers to give them a grounding in the basics of implementing artificial intelligence. The idea is that this will make it possible for them to deploy intelligent models in the company’s products.

LinkedIn has a great need for artificial intelligence, since its business is built on the back of recommendations. The professional social network recommends potential connections, jobs, content, marketing opportunities and a wide variety of other interactions.

“AI is like oxygen at LinkedIn, it permeates every single member experience,” Agarwal said. “And just to give you an idea of the scale, we process more than 2PB of data both nearline and offline every single day.”

The academy isn’t designed to give engineers an academic grounding in machine learning as a general discipline. Rather, it’s intended to prepare them to use AI in much the way that they’d use a system like QuickSort, an algorithm for sorting data that’s fed into it. Users don’t have to understand how the underlying system works, they just need to know the right way to implement it.

That’s the goal for LinkedIn, Agarwal said. Thus far, six engineers have made it through the AI academy and are deploying machine learning models in production as a result of what they learned. The educational program still has a ways to go (Agarwal said he’d grade it about a C+ at the moment) but it has the potential to drastically affect LinkedIn’s business.

Agarwal’s comments echo — to a degree — those made earlier in the day by Gil Arditi, the product lead for Lyft’s AI platform team. In Arditi’s view, one of the most pressing problems facing the ride hailing company is a dearth of qualified AI talent.

As it stands, creating a machine learning system still requires a number of skilled practitioners who can help wrangle the data and tweak the parameters needed to optimize the system. Those employees are generally hard to come by and expensive to hire. Distributing AI knowledge across LinkedIn’s entire organization could help the company keep up with its demand for intelligent capabilities.