As I reflect on how the wireless landscape has changed in the past 15 years since I was an early member at a Wi-Fi startup, what is clear is that the pace of tech innovation is accelerating with mobility and the cloud. When all of the incumbent vendors built their architectures, things like the smartphone (iPhone, Android), tablets, built-in Wi-Fi, the Internet of Things (IoT) and the modern cloud didn’t even exist yet. Today in most enterprises, people already carry 3 or more connected devices (smartphone, tablet, watch, laptop, etc.). And now, Internet of Things (IoT) devices (connected lights, displays, HVAC controls, sensors, etc.) are entering in the workplace in soaring numbers, with new devices being created and deployed daily.
In order to keep up with the soaring numbers of devices and apps, you can’t keep doing what you have always been doing. You need to change the way you: (a) connect your devices, and, (b) manage your network to deliver a robust user experience.
Your IT team needs an AI-driven assistant. You need the Mist Virtual Network Assistant (VNA). It is now here. Today.
What is VNA?
The Marvis Virtual Network Assistant is like having a Wi-Fi expert in your back pocket. It is a Natural Language Processing (NLP) AI-driven assistant that can help your IT staff to troubleshoot network issues. It can even provide you rich insights into how your network is operating and being used. Think of it as an “Alexa” for IT. It is built upon Mist’s Marvis AI engine and leverages our rich data and our data science toolkit in order to comb through the 140+ client states and make accurate conclusions of the root cause and scope of impact when issues happen.
Gone are the days when you need to have an engineer drive to one of your sites with a laptop wireless sniffer and painstakingly try to reproduce an issue for days. And no more dashboard and CLI hunting to collect data and then analyze it with your years of wireless and networking expertise. Now you can bring that expert analysis within reach of any of your IT generalists.
What are the key tools that VNA uses?
VNA takes our AI foundation in data collection, classification and analysis and applies new tools such as:
Bayesian Inference – identifies causes with the highest probability of association to the problem occurring on the network, resulting in more accurate root cause analysis.
Mutual Information – used to determine the scope of impact of an issue (e.g. just one user, a group of users, everyone at a site, etc.) by correlating performance to SLE metrics and ranking their impact.
Natural Language Processing – making the VNA easily accessible to IT staff by enabling them to ask written language questions of the system and get responses in an easily understood fashion. Now you don’t need to remember which CLI command or where that RF dashboard is located to troubleshoot an issue or find out how many iPhone devices you have on your network that are still on an old, vulnerable OS version…just ask Marvis.
How well does it work?
With all my years of wireless experience, it never ceases to amaze me how spot-on the Marvis Virtual Network Assistant’s root cause analyses are. Years ago, I remember being sent with a packet analyzer to fly from San Jose to Tokyo to troubleshoot issues for a marquis customer at a prior company. Now we have taken the Wi-Fi and network troubleshooting expertise from our experts and taught Marvis how to troubleshoot issues and put that expertise in everyone’s hands. And Marvis is learning all the time as it ingests more metadata and observes new problems. That is the power of machine learning and global metadata analysis. We started alpha testing in the second half of 2017 and deployed the beta version of VNA at several of our Fortune 500 customers in the first half of 2018. That is over 6 months of learning and counting.
Today our Virtual Network Assistant cloud subscription service is generally available to customers worldwide.
To see the Mist VNA in action, check out this short demo video that our team put together:
While this is exciting to see where VNA is today, what really excites me is that we are only just starting to see what is possible as Marvis learns over time and VNA becomes smart, both in terms of getting better and better at troubleshooting but also evolving itself to ultimately fix problems for you, without your IT team needing to take action. One day I foresee that VNA will identify the root cause as a coverage hole in a tricky spot in the office, and then it will open a ticket to place an order for an additional Wi-Fi AP, waiting for your approval on the purchase order and providing you with a justification letter that you can review and approve. Your IT team can now focus on more strategic projects, aided by Marvis’ Virtual Network Assistant.
To see more, check out our on-demand Mist Demo that also includes our Virtual Network Assistant.
Watch Now
Mist…AI is in the Air.
Christian Gilby
Director of Product Marketing, Mist Systems
Twitter: @AI_for_IT