Out of the Mist: How Juniper is leaning on AI to win enterprise market share

  • In Q2, Juniper reported that its AI-driven enterprise segment’s revenue increased 63% year-over-year.

  • Juniper added Mist AI to its network access control (NAC)  platform enabling it to natively integrate NAC into networking and security services.

  • The company views AI as an enticing opportunity to make more inroads into the enterprise market, and take market share away from its competitors.

Juniper Networks is reaping the rewards of its artificial intelligence (AI) business and working to integrate the technology across its entire portfolio. As a result, the vendor is able to use data collected from cloud applications to build models that can predict how certain applications will perform on the network, which in turn, helps with the customer experience.

In the second quarter, Juniper reported that its AI-driven enterprise segment’s revenue increased 63% year-over-year. Plus, Juniper’s Mist AI business grew by nearly 100% year-over-year and its orders increased by nearly 40% year over year.

Mist Systems, which Juniper acquired in 2019 for $405 million, was initially intended to strengthen Juniper’s Contrail software-defined wide area network (SD-WAN) offering by combining it with Mist’s WLAN platform. But with AI’s growing momentum, Juniper is capitalizing on the trend and extending Mist AI across its entire enterprise business. 

For example, earlier this year Juniper added Mist AI to its network access control (NAC) platform enabling it to natively integrate NAC into networking and security services.

Juniper said this integration allows it to better troubleshoot and resolve any network issues and also predict how certain applications will perform on the network.

“Not only can we predict the customer experience, we can figure out where the problem is,” said Bob Friday, Juniper’s chief AI officer, CTO of the company’s enterprise business and the co-founder of Mist Systems. Friday added that the company can determine if the problem is in the device, or at the service provider switch, or somewhere else.

He said that when Juniper purchased Mist the company was focused on adding AI to the edge of the network where the access point resides. However, now it extending AI beyond the network edge to the switch and the router.

“We are going to the full stack,” Friday said.

Juniper has had success with its Mist AI with certain enterprise verticals such as retail, healthcare, higher education and finance. Friday said that industries with lots of branch locations tend to want to use the cloud and have fewer IT resources, particularly at their remote locations, so the company’s AI solution is particularly appealing.

Plus, enterprises are much more comfortable using AI in their networks, making it an area of opportunity for Juniper, which currently competes for enterprise business along with companies like Cisco, Arista and Aruba. “We believe there is a growth opportunity in the enterprise market,” Friday said.

However, the service provider market is a much tougher sell. Friday said that because service providers have bigger routers and more complex networks, they are much more reluctant to have anything touch their data plane, which is the part of the network that carries subscriber data. “Service providers are comfortable sending telemetry data to the cloud but not their data plane,” he said.

Indeed, during Juniper’s Q2 earnings call with investors the company said that it experienced weaker orders from its service provider customers. The service provider business in Q2 accounted for $474 million, up 1% year-over-year but down 14% since Q1. By comparison, Juniper’s enterprise business accounted for $646 million, a 38% increase year-over-year.

Out of the Mist

Now that AI is being used to connect devices to the cloud and send workloads back and forth to the cloud, Juniper’s next goal is to use AI for real-time operations. “We are trying to build a solution that can manage networks on par with IT,” Friday said. “We are taking it to the next level.”

But taking AI to the next level isn’t an easy task because often it requires an architectural change to the network, and an organizational change, which is something most companies struggle with. For example, Friday said that when AI gets moved into other areas of the network data science team members often have to act in conjunction with customer support team members.

Nevertheless, Juniper views AI as an enticing opportunity for the company to make more inroads in the enterprise and take market share away from its competitors.