Amazon Web Services (AWS) and Lumen Technologies commercialized the AWS Interconnect – last mile service, providing a simplified approach for enterprises to configure fiber connectivity toward making the “plumbing” of cloud networking “boring.”

The AWS Interconnect – last-mile service was initially unveiled at the hyperscaler’s re:Invent show late last year. The service allows users to connect branch offices, data centers, and remote locations to AWS using Lumen’s Cloud Interconnect platform and AWS Console at a price tied to performance needs, scalable from 1Gbps up to 100Gbps.

The service is now generally available with Lumen as the first named program partner. This provides potential customers with access to Lumen’s reach of more than 340,000 route miles.

Rob Kennedy, VP of network at AWS, explained to SDxCentral in an interview that the preview period allowed for fine-tuning of the service.

“Most of the work from that preview is all the customer experience stuff, full API implementation, and then putting in the relevant capacity availability in multiple locations as well, where we interconnect with Lumen,” Kennedy said. “It's really just making sure the full end-to-end service is fully ready. Whereas the gated preview we did at re:Invent, it was still a clunky process for people to get their stuff turned up, whereas now, it's API driven, in minutes you can turn yourself up, and away you go.”

Scott Yow, SVP of product at Lumen, further clarified those last few months of work.

“It was kind of clunky where we just kind of send emails back and forth, and now we're actually doing much more deep API integration, and that's ultimately where we're driving to,” Yow said, adding that this work aligns with Lumen CEO Kate Johnson’s focus on “cloudifying telecom, and this is a big part of trying to allow the network and the boundary between network and cloud, and how you connect to the cloud, to really make that a lot more of an integrated experience.”

Kennedy noted that the platform’s API development is shared with AWS’ Interconnect – multicloud offering that was unveiled alongside last mile and that is also now generally available. The multicloud iteration is a private layer-three connectivity service that links Amazon Virtual Private Cloud (VPC) directly to VPCs of other cloud providers. Google Cloud is the initial multicloud launch partner, with Microsoft Azure set to be added later this year.

“This API standard … we developed it with an open standard mindset, so we wanted to put this out there,” Kennedy said. “We want everybody to be able to do this the right way, and it doesn't matter … [we] could do it with whoever, we'll do it with other providers as well. This just hopefully transforms the industry into a place where it is just simple and the plumbing is just – API calls or clicks of a button – makes it all boring.”

Pricing, performance, and positioning

With that “clunky” interface now excised, pricing simplicity was also touted.

“There's just a bandwidth charge,” Kennedy explained. “Do you want one-gig? We charge you just for one-gig, and transfer as much data as you want over that one-gig circuit. I think for a lot of customers in this area, they prefer that type of model. Obviously, the pay-as-you-go model works very well inside the cloud in terms of, ‘hey, if I'm not using any resources, I don't want to pay for it.’ There are pros and cons of both. But, I think what we heard from customers is that they prefer a single static charge every month. So that's what we're offering them.”

Kennedy added that while performance can range up to 100Gbps, the platform actually provisions considerably higher throughput, so it can support service level agreements (SLAs).

“If we sell 100-gig to somebody, it's actually 400-gig on our side ... which means, I could take a router out no matter what, plus I can handle two more failures overlapping and still provide that bandwidth,” Kennedy explained. “That's where we get comfortable with providing a high SLA with it as well.”

Yow noted this pricing and performance approach should ease deployment concerns.

“Historically, customers would have to think about, ‘how do I want my topology to look? How do I create redundancy? How do I deal with complex, routing, policy, segmentation, all the things that go along with that?’ And we're just dealing with all that for the customers on the back end,” Yow said. “They don't really have to think about that anymore. You don't have to engineer it, if you will. They can just consume the capacity.”

Lumen, during its most recent earnings call, reported that it ended last year with 17 million miles of “intercity” fiber assets, with company management telling investors that it plans to extend that footprint to 58 million miles by the end of 2031. Yow explained that number represents the total length of fiber Lumen has deployed across its route miles.

That footprint is currently being utilized at 72 percent of total capacity, with hyperscalers noted as using 50 percent of that total capacity and the rest being used by enterprise channels and Lumen’s own services. That overall utilization is expected to grow only slightly as the footprint expands, but hyperscalers are forecast to increase their utilization to 59 percent of that total 2031 capacity.

As for where customers for the new service might come, Kennedy said the new offering could help consolidate that decision-making.

“Things like Direct Connect, and some of our customers have hundreds of terabits worth of capacity via Direct Connect, those will continue on because they are very custom setups,” Kennedy said. “But if you think about the vast majority of customers, we think that they're going to move to this much simpler, managed option. And then we can give them all of the types of connectivity that they want, whether it's private last mile, whether it's public, whether it's satellite, whether it's to another cloud; you shouldn't have to care about where it is. It should be available at the click of a button.”

Lumen looks to maintain simplicity

This mindset echoes Lumen's long-standing move toward cloudifying access to its physical network. As Yow mentioned, Lumen CEO Kate Johnson has been vocal in describing the networking vendor’s importance to the broader ecosystem, especially as it looks to support surging AI-related traffic.

“The AI economy needs a programmable network,” Johnson told SDxCentral during a recent interview at this year’s Mobile World Congress event. “The network has become the supply chain for data, and it needs to be agile and secure and high bandwidth, and it needs to be connected to everything so you can move data from anywhere to anywhere.”

Johnson added that Lumen has constructed an architecture allowing it to take advantage of this opportunity, an architecture Yow said is highlighted by the AWS arrangement.

“It really does form the foundation for how I think a lot of the cloud on ramp should evolve over time,” Yow said. “We would love to be able to replicate this as broadly as we can. I think it's better for the industry overall. It's really kind of taking this cloud model of automated provisioning on-demand services and bringing that to the network.”