The Amazon Web Services (AWS) re:Invent event was bigger than ever this year, with 50,000 people at the in-person shindig in Las Vegas, and about 300,000 more online, taking in more than 2,300 sessions. Massive! I attended in person, and being there felt like coming together to create a small city at the center of the public cloud universe. It was thrilling to see everyone talking about the public cloud, how to use the AWS stack, and what’s next.

You could feel the excitement in the room on Tuesday morning when AWS CEO Adam Selipsky took the stage to deliver his keynote. Despite criticism that the announcements lacked luster—more integrations and fine-tuning than game-changing new products—there are still plenty of things for telcos to get excited about.

My biggest observation is the role that our industry played in the conference. My re:Invent preview blog lists some of the sessions telcos presented about how they’re using AWS. We heard from DISH, Spark, Swisscom, Telenet, Verizon, and Vodafone. We also heard about partnerships with a division of Deutsche Telekom, Telenor and Telefónica. We’ve never seen this level of participation from telco before, and I expect it will only grow over the next few years. YAY!

Beyond that, AWS announced a lot of technology improvements that are good news for telcos. Here are the best bits.

**Data and analytics FTW **

The night before his keynote, Selipsky tweeted that going into re:Invent, we should look for “the end-to-end data story.” It wasn’t hard to find, nor that surprising as he is the ex-CEO of data analytics firm Tableau (acquired by Salesforce in 2019). Data was a major focus all week. The amount of data organizations create and need to manage is growing across industries, which is especially true in telecommunications. To manage it right, Selipsky says, you need the scale of the AWS cloud. And he put his money where his mouth is, making a slew of announcements about products and integrations that make it easier to organize, manage, and share data.

I particularly enjoyed hearing from Cedrik Neike, CEO of Digital Industries, Siemens. His organization is working to help industrial clients wrestle the terabytes of data they produce every month into something useful, like business insight to help with innovation. Telcos should listen to this use case; what he described on-stage is exactly what we’re building for telco with Totogi, where I’m acting CEO, using all the same AWS tools.

Data warehousing: Amazon Redshift

AWS made a bunch of improvements to its data warehouse, Amazon Redshift, to simplify how you can get data into it, and ensure better security and reliability once it’s there, so you can put it to good use. A good first step in moving to the public cloud is moving your data up and into the cloud, and we are seeing several telcos around the world starting with this as a first project.

A data lake service: Amazon Security Lake

A data “lake” is a repository where you can keep your structured and unstructured data, and access it for processing, analytics, and machine learning. The newly announced Security Lake automates the management of security data specifically, supporting quick responses. And it can wrangle on-premise, AWS, and third-party data. The preview is available in three regions in the US, two in Asia Pacific, and two in Europe.

Amazon DataZone

The new DataZone is a data management service that helps organizations catalog, share, and govern data. Like Security Lake, it operates across AWS, on-premise, and third-party sources. The idea is to provide a single service that “sets data free,” balancing strong governance controls with streamlined access to make it easy for every employee to find and organize data, and put it to use.

For sensitive data: AWS Clean Rooms

Concerns about safeguarding sensitive data may be one of the stumbling blocks keeping telcos from using the public cloud. The new AWS Clean Rooms can help you analyze datasets and collaborate with other companies more securely, without revealing underlying data. This is brand new and still in preview. I’m super curious about it because it seems like it will really simplify things as telcos share data with vendors like Totogi to help them gain deeper, more valuable insights into their data.

And that’s not even half of it! For a complete list of the announcements about analytics (and everything else), see this blog by Jeff Barr, Vice President and Chief Evangelist at AWS (whom I met by happenstance! How amazing was that!).

Chips ah** oy **

AWS’s chips are a huge differentiator among the Big Three hyperscalers. No other cloud comes close to these hardware innovations.

The preview of Graviton3 was announced at last year’s re:Invent, and it’s now generally available. It’s 25% faster than Graviton2, which was already the fastest chip around. Graviton chips power a bunch of AWS compute services:

• Amazon Elastic Compute Cloud (EC2), which has some newly announced Inf2 instances in preview and several more types are in the works, with each instance designed for a specific kind of workload
• AWS Lambda (now with SnapStart)
• AWS Fargate
• Amazon ElastiCache
• Amazon OpenSearch Serverless (just announced and in preview)
• Amazon Aurora

The Nitro System is powering the next generation of EC2 instances, where AWS is continuing to innovate for even faster, better, and cheaper performance.

AWS’ custom chips are one of its strong differentiating features over the other hyperscalers. It’s something AWS invested in years ago, and one way customers can save a lot of money by reducing their overall compute needs. NTT Docomo has discovered this already and wrote up findings for its 5G Core where it saw a reduction of 72% in power consumption; read about the results in this announcement.

**Shout out to call centers **

There’s huge potential to improve call centers, and customer experiences in general, through artificial intelligence (AI) and machine learning (ML). It’s one of the main areas where telco data can do the most good. Amazon Connect, a “cloud contact center,” has two new ML-powered updates in general availability, and two more new capabilities in preview.

Amazon Connect already makes it easy to set up a contact center in the cloud, analyze voice conversations in real time, help agents find relevant information, and authenticate customers by the sound of their voice. The new stuff:

• Forecasting, capacity planning, and scheduling (generally available)
• Conversational analytics for chat (generally available)
• Automatically scored evaluation forms for agent performance (in preview)
• A step-by-step experience to guide agents through resolving issues (in preview)

There’s a lot of room for improvement in contact centers, and a lot of opportunity to do it with better data analytics, AI, and ML. I often say that the main benefit of moving to the public cloud is all the services you get access to. Contact center software that keeps getting better via cutting-edge innovations from the world’s best technologists is a great example, and since telcos still have some of the largest call center organizations in the world, maybe this is something for them to watch more closely. By adopting the AWS contact center technology, you can rapidly add more features, faster than you can with your existing on-premise legacy system. Avaya, watch out.

A few final observations …

Sustainability was another theme at this year’s event. Lots of telcos are setting goals around reducing their carbon footprint and reporting on their progress annually. Public cloud vendors have been getting a bad rap for years about the amount of energy they consume, so it’s nice to see such a strong focus on efficiency and on reducing consumption and waste. For enterprises, the public cloud’s elasticity can really help in this regard, as it allows you to scale up and down to use only what you need at any given time.

Also, IMHO, Adam Selipsky had a big miss when he glossed over growing concern around managing costs. He touched on it briefly with a story about Airbnb, but especially with the recent high-profile repatriations to on-premises, he missed an opportunity to share how much Amazon really wants you to save money with this move. I’d love to see all of the hyperscalers really help customers to manage their cloud spend with best practices; when you do it right, you can reduce your spend by 50% each year.

But all that aside, it’s clear that AWS is still the leader among hyperscalers, largely due to its head start and massive scale. It now has 30 launched regions in 21 countries (plus five more in progress) and 96 availability zones (with 15 more in progress). It offers more than 200 services that it constantly improves, and offers over 100,000 APIs. All in all, AWS is hard to beat.