Deploying AI Models at the Edge

An interview with Kurt Daniel, CEO, Ubersmith

An interview with Kurt Daniel, CEO, Ubersmith

Edge Industry Review has attended many edge computing-related events over the last year. After listening to speakers at Edge Congress America, we recently caught up with Kurt Daniel, CEO of Ubersmith, a provider of subscription management and related software for service providers. Ubersmith is a company that is positioned to see what kinds of business models are emerging for service providers in the edge computing market, so we asked Daniel via phone and email about his thoughts on what’s driving the market and his views on the market’s evolution in 2020.

Edge IR: In your view, what is driving interest in edge computing, and what are the drivers for growth?

Edge computing is a great name for an exciting trend that builds from the strong shift over the past few decades to the cloud, SaaS and third-party data centers. The edge can be viewed both in a broad context meaning getting computing and storage closer to the user and business or in a narrower context via specialized use cases such as IoT on-site to name one prominent example.

I see the following future drivers for edge computing:
• The edge complements the cloud while the cloud dwarfs the edge
• Highly specialized edge use cases prove early success
• Cloud players build out further around the world and claim closer data centers and cloud locations as edge
• Broader edge use cases prove successful after the underlying technology infrastructure and management solutions are ready for prime time
• Leading-edge vendors are made obvious by the market

The promise of the edge in some sense is a shift back in part from the cloud and centralized infrastructure toward on-premise computing.

Edge IR: What markets will be key to driving edge computing expansion in the U.S. and beyond?

I think about the IoT use case first, including the oil and gas local use case. Therefore, I would say first any location, regardless of region, where this work is done and where the cloud is too slow, too expensive or otherwise not ideal. I also think a lot about manufacturing and industrial IoT throughout the U.S.

Additional locations come into play based on additional IoT and other specialized use cases, including mobile health, or with a broader definition of edge computing which I highlighted above.

Edge IR: Apart from the U.S. market, do you have any other regions where you feel edge computing could accelerate in 2020?

We’re currently working with organizations on six continents in general across our business. I’ve seen the most interest in edge computing in North America and Europe, but this may be a function of us being based in New York City. I can see edge being helpful everywhere.

Edge IR: Will IoT also be the key use case in other regions as well, or would it be something else such as online gaming?

We think both of those use cases – IoT (for residential, commercial and industrial use) and gaming (especially with the new game streaming platforms coming in 2020 from Google and others) will be popular use cases. CDNs are another key use case.

Edge IR: Would there be any insights that an edge infrastructure service provider might glean from SaaS providers or other types of customers you have? Any trends you see that they should take note of?

I haven’t seen others talk about it, but I think SaaS can be a tremendous use case for edge computing the more I think about it. I think it’s more about reaching higher levels of performance, analytics, and deep learning anywhere in the world than unique business models for SaaS companies and the edge. I’m excited to see other trends emerge also beyond the obvious ones we’ve all read about (oil and gas platform, gaming, etc).

Edge IR: Maybe the trend or issue is more about the device integration – I see a broader array of CPUs and GPUs, FPGAs and now AI-specific chips being used for edge workloads. Do you have any thoughts to share on enabling faster time to revenue by leveraging your integrations?

Our hardware and software integrations are usually not touching the chip level for example, but we do want to see where things are going and offer new integrations in device management when possible. We do see in edge computing how running device management agents locally in data centers can improve performance, for example.

Edge IR: Can you talk about some of your strategic goals for the next year and how edge computing factors in?

One of our big goals at Ubersmith is building upon our strength of integrations tied to our billing, ticketing and related software for data centers, cloud providers and others. We currently offer 250+ software integrations / supported devices which enable turn-key billing for virtually any type of computing resource as well as location-based device management of similar resources.

We are excited to offer new hyperscale integrations. We started with AWS and are in final testing with Azure. As we look to extend this to additional providers, we are interested in seeing them offer edge computing services that we can enable from our suite of billing, infrastructure, and operations functionality while looking to partner with pure edge solution providers and others as well.

About the subject

Kurt Daniel has led Ubersmith, a provider of subscription management and related software for the cloud and beyond, since 2014. Previously, he helped build and scale three software startups – Parallels, Worklight (acquired by IBM) and MongoDB – into companies worth $10 billion. Earlier, he was a Lead product manager at Microsoft. Kurt is an independent director, serves as a venture partner at NextGen Venture Partners and mentors startup founders at Antler Global. He has a Bachelor of Economics from UVA and a Master of Business Administration from Harvard Business School.

Article Topics

 |   |   |   |   |   |   | 

Comments

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Featured Edge Computing Company

Edge Ecosystem Videos

Automating the Edge

“Automating

Deploying AI Models at the Edge

“Deploying

Latest News