In March 2019, Juniper Networks announced it was acquiring Mist Systems, one of the pioneers in cloud-managed WiFi. While there were other cloud WiFi vendors in the mix, what made Mist stand out was that its solution is powered by artificial intelligence (AI), emphasizing the importance of AI in networking.
At the time of the acquisition, Mist was used to improve the performance of Juniper’s WiFi solution. Since then, Mist’s AI-based management has been applied to the company’s campus networking products and is soon coming to SD-WAN and data center.
Mist is the foundation of Juniper’s enterprise strategy
To get a better understanding of how Juniper will be expanding the use of Mist and the value it brings customers, I recently interviewed Mist co-founder and CTO, Bob Friday, on a ZKast video interview, done in conjunction with eWEEK. I encourage everyone interested in how AI will impact network operations to watch the video, but I have provided highlights here:
- Under CEO Rami Rahim, Juniper has doubled down on enterprise. He expects the enterprise business to soon be larger than Juniper’s business for service providers, although he provided no timeline for the changeover.
- Mist was a key acquisition for Juniper’s enterprise business, because it’s become the foundation for its networking services. Since then, Juniper has made two more key acquisitions: 128 Technology in the SD-WAN space and Apstra in the data center. The three companies combined can bring fully automated operations across all aspects of the business network.
- Mist was designed from the ground up to leverage the elasticity and scalability of the cloud in order to simplify networking through the use of AI.
- The cloud is fundamentally a better way to maintain software. It’s why virtually every other area of IT has migrated to a cloud-first model. The network is the last area to leverage the power of the cloud, and AI will drive this transition.
- The goal of Mist is to have the troubleshooting capabilities of a high-level network engineer.
- Although there are many vendors that claim their systems are AI-based, a number of them are actually rules-based products. The best way to identify true AI systems is to see if the product improves over time. AI platforms are constantly learning and should get better with more data.
- Marvis, Mist’s Virtual Network Assistant, has been increasing over time and is now at 70% efficacy (see chart at 8:20 of video). At 90% it will be on par with a championship-level network engineer.
- Marvis / Mist is not designed to replace network engineers. The system should be looked at as a tool that can help network professionals do their jobs more efficiently. ZK Research shows that about a quarter of network engineers spend about one day a week doing nothing but WiFi troubleshooting. Mist can remove much of this heavy lifting and enable engineers to fix problems instead of diagnosing them.
- Marvis provides a fundamentally different and more efficient way of running a network when compared to the traditionally manually intensive command line interface.
Zeus Kerravala is an eWEEK regular contributor and the founder and principal analyst with ZK Research. He spent 10 years at Yankee Group and prior to that held a number of corporate IT positions. Kerravala is considered one of the top 10 IT analysts in the world by Apollo Research, which evaluated 3,960 technology analysts and their individual press coverage metrics.