Real-Time Data at Verizon: It’s as Critical as Air

By Reed Peterson, Field CTO, Telecom, DataStax

Efficiency, speed, growth, and excellent user experiences. These are goals that digital enterprises strive for regardless of industry segment. In the hyper-competitive telecommunications market, companies that don’t achieve these superlatives risk being left in the dust by competitors.

It’s one thing to deliver performance and new services that impress customers, but doing so at the scale of one of the largest wireless carriers in the world and at the speed required to compete in a cutthroat industry requires data – petabytes of it. As well as a clear understanding of how to take full advantage of it in real-time.

Greg Sly, Verizon’s senior vice president of infrastructure and platform services, oversees storage, compute, and networks related to data centers that serve Verizon, along with a set of CIO-level responsibilities, including supporting all the technology used by the company’s employees to run its corporate offices and retail operations. We recently spoke with Sly about the challenges data creates for big organizations and the importance of real-time data.

What’s the biggest challenge data creates for a large enterprise?

The biggest challenge for any big enterprise is organizing the data that has organically grown across the organization over the last several years. This isn’t unique to Verizon. Everyone has data lakes, data ponds – whatever you want to call them. They have all grown up organically within various business units. Now it’s about bringing that together. How do you get your arms around all the data you have?

Verizon is one of the largest telcos in the world. Imagine the size of some of the data stores we have. Any one of our business units is almost a Fortune 100 company on its own. When you pull all that together, you’ve got a lot of complexity with a lot of data that needs to get categorized, inventoried, and controlled.

How do you solve this data complexity?

We have to find out what we’ve got, and where we’ve got it. And then it’s really choosing the right architecture and tools. What’s the platform going to be going forward? How do you migrate from your current, multiple implementations with different tools into a centralized platform? You need an enterprise solution that’s scalable, usable by everybody, cloud-agnostic, and doesn’t require everyone to do a complete code rewrite of everything they’ve done up to this point.

How important is real-time data in building innovative customer experiences?

Real-time data is air. Real-time data is critical for everything, from your financials to customer sales to whether promos are working, whether advertising is working, whether the initiatives that you’re investing in are paying off. You’ve got to pivot so fast in today’s market. If you don’t have real-time data, you’re back in 1985.

Real-time data is not a nice-to-have anymore. If you can’t make real-time decisions, and if your experiments are running for three months, forget it, your competitors will have gone on to the next thing and you’re playing catch up. 

One of the critical enablers is our ability to pull in feeds from different areas into multi-streaming messaging and then make real-time decisions. You need to be able to give the CEO, the president, or the business the real-time information to make business decisions for the company. So, real-time data has become air.

What role does Apache Pulsar play in Verizon’s data architecture?

We’re using a lot in the consumer side of the business. That’s probably our biggest implementation right now: event streaming feeds a lot of data back into our customer service and consumer data workflows. Pulsar is also making more inroads on the enterprise and business group sides. We’re starting to use a little bit more on the infrastructure side as well. We’re using it a lot with call centers and event streaming and customer service and the tools that we do for provisioning.

How important is open source software and the innovation it provides?

Open source is very important, and I believe it’s a key way to accelerate innovation. The open-source community has already solved some of the hard problems we face, and they’ve done so with really innovative solutions. Depending on the technology, there can be thousands of developers out there either with a similar problem or a way to a solution that they’re more than happy to share –  and that we can participate in building rather than building from scratch. Any company is crazy not to try to leverage the open-source development community because there’s a lot of geniuses out there who love to share their knowledge. Companies should keep in mind, though, that the open source community is a give-and-take. So if they plan on adopting open source, they should also be contributing to the community. That’s what makes it what it is: the balance, the expertise, the community.

How important is data to innovation for your internal developer teams?

We’re trying to give our developers access to larger data sets when it will enable them to make better decisions on what’s going on in the market, what’s going on with customer asks, what’s going on with testing for QA. The cleaner the data set, the better the product and experience is going to be.

With developer access to more, appropriate, data, we’re already seeing higher ratings on the App Store, we’re seeing fewer bugs, and we’re seeing fewer customer complaints. We’ve been looking at our service tickets and our call volumes and they’re continuing to go down because we’ve got better data to make better decisions and then better testing and regression testing because we have better datasets to regression test against. We’re getting a better quality product out the door, which allows us to innovate faster because now you don’t have 25% of the team fixing bugs from the code we deployed last week.

How does DataStax enable Verizon?

We transitioned from batch processing to streaming and in doing so have been able to reduce our data services hardware budget by more than 30%. DataStax is one of the core technologies that helped us achieve that goal.

DataStax actually paid for itself several times over, just on my hardware budget. We realized we can start doing data faster, and we can start consolidating down to a single platform. And the nice thing was, we didn’t have to go back to the development or product teams and ask them to rewrite all the code. DataStax has the APIs and the pipelines to just say, “Throw it on the message bus, get it over here. We can deal with it – and hey, by the way, if you want to do some tuning later, that’s fine, but you don’t have to do a single thing.”

How important is the working partnership with DataStax?

A great thing about working with DataStax is we got to engage with really smart technical people early on: people who had done this, they’d been in the trenches. So there was a lot of credibility with our engineering staff early on. The fact that we could get architects and the senior people from data warehousing in the room with senior DataStax guys who knew their stuff made a big difference. We don’t normally get that with other vendors.

Learn more about DataStax here.

About Reed Peterson:

Reed leads telecom strategy and engagement at DataStax, focusing on digital transformation and new market opportunities in the 5G, IoT, Edge and AI/ML spaces. He was previously at GSMA,the global trade association that represents the mobile industry, and he spent a decade in management and strategy consulting, focused mostly setting up US companies in global markets.