“Is it really AI?”
We get asked this question all the time, and it’s a fair one. There is quite a bit of “AI-washing” in the industry as vendors with the slightest bit of automation are quick to apply this moniker to their products.
To answer this question, let’s look at the simplest definition of AI – when you can’t tell the difference between a human and a machine. This cannot be achieved with just automation; it also requires insight and learning. For a machine to be indistinguishable (or even better) than a human, it needs to understand the actual intent behind a question and choose the best answer for solving it. This is not always so easy – just look at the IBM Watson/Jeopardy experiment to see how homonyms and plays on words can wreak havoc on this process.
Mist has always been on a journey to create a WLAN platform that could answer questions on par with an RF expert. Similarly, both Mist and Juniper Networks (Mist’s parent company) have had a mutual long-term goal of delivering self-driving networks, where self-optimization and problem resolution are both automated and accurate. 
Today, we announced two new enhancements to the Mist-Juniper portfolio that further these objectives:
The first is Marvis Actions, a new capability within the Mist Virtual Network Assistant Service that leverages our integrated AI-engine (Marvis) to not only identify the root cause of issues across various IT domains (WLAN, LAN, WAN and Security), but to automatically resolve them when possible, such as automatically adding a missing VLAN tag or fixing an incorrect port mode configuration setting, or to provide prescriptive recommendations on how to fix them when the issue is found to be with an external system. It is the first, and in some instances only, place our customers have to look to monitor and manage the health of their Mist network and brings unique self-driving and driver-assist capabilities to IT teams.
The second game changing innovation we announced today is a new Wired Assurance service, which extends the Mist’s AI framework and unique Service Level Expectations (SLEs) to Juniper’s EX series of access and aggregation switches with their rich telemetry that Junos provides in order to deliver better network experiences for wired devices such as IoT devices, access points, servers, machinery, printers and beyond. This brings a new level of automation and insight to the wired world, including anomaly detection, switch health metrics, and the Marvis Virtual Network Assistant (with Marvis Actions) for easy troubleshooting, insights and integrated help-desk functions. While many vendors simply focus on management capabilities of a unified portfolio, we are the only vendor who is transforming IT operations with AI and insights across the network stack and beyond.
With these two announcements, Juniper is delivering on a major goal of the Mist acquisition – to unify the various components of the access network under a single cloud and AI-engine which will help customers transform to an AI-Driven Enterprise. For our customers, this means more visibility into user experiences coupled with even broader automation for more seamless and cost-effective IT operations. In addition, we are now two steps closer to bringing the industry true self-driving networks, which is the ultimate example of engineering simplicity at work!