Ms. Logothetis is executive group owner of edge, cloud and industrial networks at Telstra, which encompasses dedicated and private networks. Her job is to accelerate the adoption of these technologies by enterprises in parallel to and stimulated by the convergence of “the best of global compute with Australia’s best connectivity.” Although she has only been in her post at Telstra since February of this year, her rounded CV has prepared her as well as any for her pioneering role in crucial new territory for Australia’s biggest telecoms service provider.
“We are really looking at automation from the aspect of how we best deliver an amazing product experience to organizations across Australia. Our automation goes all the way from our network through our IT stack and to the way we build strategic partnerships, including with cloud hyperscalers. We want automation around all those capabilities and to make them modular and expose them through APIs,” Logothetis states.
She stresses the importance of decoupling the architecture “to build a highly digital, highly automated, amazing product experience using those capabilities. This enables us to be intuitive and agile in what we can deliver to the market; quarter after quarter, we’re getting more and more product experience out to the market. The [decoupling approach] gives us a high degree of reuse: I can reuse the capabilities in edge products. I’ve used them in cloud products and private network products. We are starting to get that consistency of experience across products as well.”
The modular, API-enabled operating model is also fundamental to collaborating efficiently and effectively with partners. Telstra works with cloud hyperscalers as its cloud compute and edge compute partners, plus some of the large data center vendors and dominant OEM-type vendors. In this ecosystem, Telstra takes on either part or all of its partners’ technology stacks in this modular way via the APIs, then figures out “how we bring that together into a product we can market to our customers,” Logothetis explains. “Luckily, I work with partners that are building technology natively in this way.”
She clarifies that “natively” in this context means both cloud native and softwarisation – building offers as-a-Service using APIs. She notes, “We are doing that inside Telstra and relying on our partners to do it so we can create this experience for the customer.”
Automation Beyond Self-Service
“People tend to think about automation as being about how a customer buys something from and engages with an organization; the order-to-activation process has always been key for telcos…Once we start to build automation, we look at the onboarding experience to make it faster and simpler for customers, so it’s just a clicking this or swipe that type of exercise.”
Logothetis further stresses, “We have very good digital portals, in the consumer and enterprise spaces, and will continue to develop them. But the more recent innovation we’re working on with our customers is understanding how they use our products and what they want, including using telemetry data.” For example, if a Telstra customer has a cloud tenancy, how can they scale it up or create a new tenancy? How can they add AI on top of it? How can they move the tenancy closer to them? In other words, “It’s less of an interaction with us through a digital channel and more of an interaction with the product itself,” she details.
While in the first instance, automating this product interaction is geared to the enterprise market where Telstra expects “really deep engagement with products,” according to Logothetis, “I think as we look out into AR and VR augmented and virtual reality, immersive experiences, and the metaverse – depending on [what] it ends up being – it becomes a very similar sort of scenario, right? It’s about how consumers engage with the product or act inside the product versus engagement between them and our organization.”
Cloud Matters
In the meantime, she says, “We talk about key things like hybrid and multi-cloud, which enable our customers to put their workloads in the place that best meets their demands. Some are best placed in the public cloud with any one of a number of public cloud vendors. For sovereignty and security reasons, some might need private cloud. For legacy reasons, some might run virtualized technology on a public or a private cloud. Other customers want private cloud workloads. We offer that spectrum.”
Telstra has multiple partners at all layers of the cloud stack, and this is where edge compute comes into play. Logothetis says, “We start to talk about distributed computing, because customers will have more and more applications and data workloads with unique sets of requirements. Some they’ll want to have much closer to them because it’s data intensive – they want to collect, store and process the data close to where it is – or because they need actionable insights from it to correct a safety issue, say, or change a manufacturing process.”
Distribute, Compute
However, these needs might only apply during the day, not overnight, when instead some applications could run at the network edge or in the cloud. Logothetis explains, “I think this notion of distributed compute, with very good connectivity between it, and some smart software sitting on top, for an enterprise or a government organization, places workloads in the right place at the right time.
“That’s what we’re working on. We already offer all those components today but where we see this industry and demand heading is that all those components work seamlessly together.”
This is not an easy undertaking. So what are the challenges in this level of automation? And are the limitations of AI an issue? Logothetis’ view is that it depends on what you’re trying to automate and why. She says, “From my role, the trickiest thing is to establish what we are trying to deliver to the customer – a better experience of something they have today or using automation to come up with a fundamentally different experience, maybe something that didn’t exist before.”
Starting With Desired Outcomes
It’s refreshing to find an organization that starts by thinking about what they want to achieve and then reverse engineers to where it is today to figure out how to get there. It’s more common for telcos to focus on a shiny new piece of technology, then figure out what they can do with it and how to justify the investment. This common wrong-headedness is often compounded by technology becoming “legacy” at the fastest rate ever.
Logothetis agrees, and furthers, “That all comes back to the principles of decoupled architecture with the APIs around it; then building this product experience we’ve talked about on top of that – that architecture and that way of interacting will enable us to be really agile in getting things to market and changing things quarter on quarter, based on what we see the customers doing with it.”
Internal Ops Enable External Experience
Telstra is also working on AIOps because, as she says, “If you have a highly manual back office or operations, it is very difficult to automate the experience. Even if it looks great on the surface, underneath, it’s like those images of swans sitting serenely on the surface, but their feet are moving frantically under the water. So internal automation and customer experience absolutely are linked, but not tightly coupled because we take a modular approach that is API-driven; then build this experience layer on top. As things change in the underlying organization or in the capabilities of ecosystem partners, we can bring things in without having to build a brand new product from scratch.”
Logothetis emphasizes how important this is and that the ecosystem is “developing very quickly, particularly around the edge cloud and in private networks. The customer-facing parts are very much an ecosystem play, because customers want a broad spectrum of capabilities, and there’s no single vendor that they want to be totally tied into. Customers want multiple different things, and a big part of our role is to work as a part of that ecosystem and make it easy for our customers to work in that ecosystem as well.”
She concludes, “We need to build the ecosystem from the experience for the customer and the engagement with the product, as well as building automation behind scenes, which is probably what you hear most about in the industry – things like automating provisioning and billing. That’s happening, but what we’re doing is really interesting – the automation of experience and interaction with products.”
A version of this article was first published in September 2022 on the FutureNet World website.
