Extreme Networks’ Chief Product and Technology Officer on AI: ‘Start. Learn. Get to a small value’
Last week, RCR Wireless News sat down with Extreme Networks’ Chief Product and Technology Officer Nabil Bukhari at the company’s annual summit in Fort Worth, Texas to find out more about how artificial intelligence (AI) is impacting the networking space and what organizations should be thinking about as they look to make the most of this gaming-changing technology.
Bukhari deviated from the typical script, which harps on the importance of becoming an AI-first or data-first company. That’s a tall order, said Bukhari, and even more, an unnecessary one. “A lot of Customers are scaring themselves … If you’re saying I cannot use AI until I change my culture, upskill everybody, clean my entire data, good luck, we’ll all be retired by then. Companies don’t move that fast. Culture doesn’t move that fast. People who are going with that mindset are going to get left behind,” he claimed.
So, what is the right mindset, then? Well, according to Bukhari, organizations should begin by dividing their functions into the following three categories:
Functions to accelerate or augment. “In networking, there are massive amounts of things that your people do that AI can just simply help them do it a little bit quicker, from troubleshooting the network to generating a report out of it,” he said, emphasizing that these should be “simple, simple functions” that, when accelerated, save a company either time or money, or at the very least “the mental health of [its] people.”
Functions to replace or avoid. These are the things that an organization no longer wants its employees to do. Bukhari provided the tedious example of generating a first draft of a policy for a new security tool. “Imagine if the AI could give you the initial draft, and you could just review it?” He proposed.
Functions to be created. “These are the things that you’ve never done before and that’s the create use cases,” said Bukhari. “These are the things that AI will enable, the things that everybody talks about. But don’t start from those create use cases because that confuses the hell out of everybody.”
When you start by first figuring out what you want to accelerate using AI, said Bukhari, the use case is very well-defined and has a very well-defined, binary success criteria: Were you successful in accelerating it or not? “You’re chances of success with these specific use cases are higher and as you succeed, you learn, so your capabilities and confidence in AI increase. Your trust increases in AI, and you are generating value on day one,” he said. “You’re creating value from the start.” And when you create value, even small value, early in the AI journey, it justifies further investment in AI down the road.
“Find the thing that you want to accelerate, figure out the one or two data sets that you need for it. Start. Learn. Get to a small value. People who go in with this iterative fashion, with specificity, I think they will get farther ahead than those with these grand gestures of I’m going to pivot my company towards AI, I’m going to change my data culture and retrain everybody,” concluded Bukhari.
Click here for more on how telco AI supports network automation and 5G monetization.
