Bob started his wireless career at Metricom (Ricochet wireless network) with the original release of the part 15 FCC unlicensed rules in developing and deploying wireless mesh networks across the country to connect the first generation of Internet browsers. Bob co-founded Airespace, a start-up focused on helping enterprises manage the flood of employees bringing unlicensed Wi-Fi technology into their businesses. Cisco acquired Airespace in 2005 for $450M. He also drove industry standards such as HS2.0 and market efforts such as Cisco’s Connected Mobile Experience. Bob left Cisco in May 2014 to co-found Mist Systems.
Mist is pioneering the new wireless network. We built the first AI-driven WLAN, which makes Wi-Fi predictable, reliable, and measurable and delivers amazing indoor location experiences.We are a unique company with strong DNA in wireless, cloud and data science. We are eager to change the world, one mobile user at a time. Are you ready for challenging work, and a lot of fun along the way?
What made you launch Mist?
I co-founded Mist with the goal of building an AI-driven wireless platform that could answer questions on par with a wireless network domain expert. Similarly, I wanted to use AI to bring indoor location experiences on par with that of outdoor GPS.
This vision for an AI-driven wireless network came to us via the following two inspirations:
In working extensively with customers, I found two glaring issues with traditional wireless networks. One, they don’t give visibility into the actual Wi-Fi user experience and troubleshooting problems is a manual and laborious task. Second, more and more companies want to move wireless beyond a medium for connectivity to a strategic platform for mobile experiences. This made me realize that a new architecture was needed to help businesses put critical services on their wireless networks and be able to innovate their infrastructure at the same speed with which they were innovating on the mobile side.
I have always been a technologist and science fiction fan. When I saw IBM Watson play and answer questions on par with some of Jeopardy’s best champions and win Jeopardy, I realized that all the technology pieces were finally available to build an AI-driven WLAN that could solve the issues above.
How does Mist leverage AI to deliver better experiences and ensure optimized operational cost?
This has been a three-year adventure so far, which continues to evolve every day. It started with building a real-time distributed cloud platform for collecting data needed to understand the mobile user experience. Then we took that data and created real-time Service Level Expectation (SLE) metrics to classify the information in a way that a wireless domain expert would understand. With classification done, we created an AI data science tool box that processes the SLE Meta data and correlates it for actionable insight. Finally, we implemented natural language processing to put a face on the AI platform, making it easy for IT to get the data it wants. The end game is a virtual AI network assistant that can answer questions on par with a wireless domain expert.
This Mist platform changes the whole IT paradigm as companies can move from managing the network to managing the end-to-end user experience (from the mobile device all the way to the internet.) This not only lowers operational costs, but it leads to better wireless experiences for mobile users.
How do you make AI deliver economic benefits as well as social goodwill?
AI-driven networks bring economic benefits and social goodwill to companies in various ways.
From a business perspective, we allow companies to move IT resources and investment into areas that are critical to their core business. For example, a retailer who is in the business of selling products can now use precious IT resources to optimize customer engagement versus wasting them on reactive troubleshooting tasks.
From a social perspective, Mist brings indoor location on par with outdoor GPS. In addition to enabling more core business services, this also can help society. For example, reliable indoor location can be tied to E911 services to save lives – e.g. has everyone been evacuated during an earthquake? Where should I send emergency responders? Beyond safety and security, the use cases for indoor location are almost limitless, and I predict will be even greater than what we saw with outdoor GPS.
Would AI close the gap between IT, and non-IT assets in an organization?
The Mist platform bridges the gap between IT and help desk personnel by giving the latter tools to easily understand the wireless network and troubleshoot problems. Help desk personnel often lack deep rooted RF expertise. But with natural language and automated event correlation, they have an IT expert sitting right next to them at all times. A virtual network assistant can quickly let customers know if their problem has been identified and scheduled to be resolved.
In addition, Mist bridges the gap between IT and business functions (like marketing) by delivering indoor location services that are simple and scalable. In the past, the person responsible for patient experiences at a hospital or guest experiences at a hotel relied on battery beacons to deliver indoor location services, such as wayfinding and proximity messaging. Now they can leverage the normal IT infrastructure for this, which ensures better reliability and saves time/money.
What does the future of wireless cloud systems look like beyond 2022?
Computing power will continue to increase and get more cost effective, especially with Quantum computing on the horizon. This will allow virtual network assistants to continue to get smarter, more accurate, and more predictive. Network infrastructures will therefore become more nimble and proactive, with self-healing and other automated tasks becoming the norm.
Tell us about your AI research programs at Mist and the most outstanding digital campaign at Mist or elsewhere?
Our main focus right now is to continue to apply AI techniques like machine learning, neural networks, and NLP to wireless networks so that Wi-Fi is predictable, reliable and measurable, and Bluetooth LE can be used seamlessly for high accuracy location services.
Beyond this, we are also working on converging network data with granular BLE location information to leverage deep learning image processing techniques. In addition, we are working on applying our Marvis AI platform beyond wireless to other domains, such as security, analytics, and IoT.
Our most outstanding campaigns come in the form of happy customers. For example, a top ecommerce company is using Mist to fully automate their wireless operations for faster and more cost effective Wi-Fi troubleshooting. Similar, a popular retail chain is using Mist to greet customers upon arrival, provide directions, find an available associate, and more. This personalized engagement gives them a competitive advantage, all thanks to Mist. And a large healthcare customer is using Mist to provide our veterans a better mobile engagement during their visits.
What are the major challenges for AI technology companies in making it more accessible to local communities? How do you overcome these challenges?
While AI and machine learning technologies are powerful and will ultimately have a major impact across all industries and society, you need to put a face on AI that allows the average person to consume them. This is where natural language will play an increasing important role. Domain experts may love their scripting and query languages, the average person does not.
What AI start-ups and labs are you keenly following?
We have been working with startups such as Ople who are focused on helping business leverage AI and datascience with fewer data scientists to solve problems.
- Cortical.io working on text base NLP
- Medial EarlySign looking at predicting a person predisposition to diseases
- Experfy on demand data science consulting
What technologies within AI and computing are you interested in?
Unsupervised machine learning to help make indoor location easier to deploy. Deep learning neural networks to help predict users mobile internet experience. New quantum computing startups such as Regetti.
As an AI leader, what industries you think would be quickest in adopting AI/ML with smooth efficiency? What are the new emerging markets for AI technology markets?
I see healthcare being one of the early adopters of AI to help doctors diagnose diseases and save lives.
Cyber security will be an early / fast adopter to keep up with the arms race against the bad guys.
The service industry will be an early adopter of AI too, as it is increasingly urgent to better engage with customers/guest and deliver a personalized experience that re-enforces brand equity.
And if I have my way, the wine industry will be an adopter of AI to reduce the number of glasses I need to drink to find the perfect glass!
What’s your smartest work related shortcut or productivity hack?
Dynamic Packet Capture … Whenever a wireless user calls with a problem from the past, I can go back and find the packets associated with that problem and figure out what happened. It is almost impossible to replicate a wireless problem so this is major time saver.
Tag the one person in the industry whose answers to these questions you would love to read:
Thank you Bob! That was fun and hope to see you back on AIthority soon.