Peacock Achieves a Fast and Flawless Launch Using AWS Solutions

The streaming service Peacock—launched on July 15, 2020, by global media and technology company Comcast Corporation (Comcast)—arrived in an era of heightened consumer demand for video streaming services. In this new era, Comcast has continued to build incredible technology and entertainment services that connect millions of people to the moments and experiences that matter most. As of December 2020, Peacock had already amassed over 26 million sign-ups.

Comcast companies NBCUniversal and Sky partnered and dedicated resources to build this competitive streaming venture and moved from idea to implementation within 12 months. Conceptualized as an over-the-top streaming service that would offer video on demand with both paid and free subscription levels, Peacock would also need the capability to support live broadcasting of events around the globe. One of NBCUniversal’s expected big-ticket live broadcasts, the 2020 Summer Olympics, determined the timeline that Peacock’s technology team would be following to launch the service. That timeline gave the team just 1 year to build an entirely new system with high availability and resiliency. Knowing it had one chance to make a great first impression on viewers, Peacock needed its launch day to be perfect. 

 In only 1 year, Peacock created a simple, derisked, scalable, and highly available system on Amazon Web Services (AWS). And despite unforeseen obstacles like the COVID-19 pandemic and the postponement of the 2020 Olympic Games, Peacock achieved a flawless launch right on time. “In the face of aggressive timelines for launch, development, and customer readiness, we coupled our resources and expertise with those of our onboarding partners and AWS to meet those goals,” says Eric Black, head of US operations and video for Peacock.

In January 2019, NBCUniversal and Sky announced they would partner for the first time to launch a joint streaming service. Peacock’s technology team envisioned a fully cloud-based system with no on-premises infrastructure. The service needed to be highly reliable at an unknown but potentially massive scale. “There are a lot of systems behind our users’ journey to their content,” says Black. “With Peacock, the infrastructure would need to be able to manage the load of a service that delivers a multidimensional experience—one that goes beyond login, search, browse, and display to provide live, on-demand, short-form, and channel-based streaming at scale.” In addition, Peacock needed to support the demands of live video streaming, which can result in aggressive spikes in viewership. 

Peacock’s workflow design had to be simple and streamlined to accommodate the tight timeline. The full workflow had to stand up to an extensive testing regime, and the Peacock team needed a team available to help analyze potential problems and correct course before the launch. These needs were key in Peacock’s choice of AWS for its cloud services. “We chose AWS for its subject-matter expertise and level of support and engagement,” says Black. “It’s not just an infrastructure or technology play; the resources and expertise AWS delivers help us make our project successful.”

Peacock dove into building its system using a suite of AWS services. The bedrock of the architecture was Amazon Elastic Compute Cloud (Amazon EC2), which provides secure, resizable compute capacity in the cloud. With a service-level agreement of 99.99 percent availability for each Amazon EC2 region, it could provide the high availability the Peacock system needed. 

While working on the network design, the Peacock team used AWS Transit Gateway to simplify and connect its virtual private clouds and on-premises networks through a central hub that acts as a cloud router in which each new connection only needs to be made once. The Peacock development team built the rest of its simple yet fully scalable system using several AWS services, including AWS AppSync, a fully managed service that helps companies develop applications faster through scalable GraphQL APIs. Peacock also used AWS Elemental MediaStore—a media storage service that provides the performance, consistency, and low latency required to deliver live streaming video content—to create video-on-demand content for broadcast and multiscreen delivery at scale. 

As the technology team built the Peacock system, AWS solutions architects helped put the system through rigorous testing. “We went through a number of AWS Well-Architected Reviews where we brought in AWS experts to review our configurations so that we would be prepared from operational, reliability, and performance perspectives,” says Stephen Hildebrand, vice president of performance and reliability at Peacock. 

While the system’s design was simple, the team had to work hard to integrate products from multiple vendors and input from stakeholders around the world. When the COVID-19 pandemic hit in spring 2020, AWS was a key player in helping the project stay on track. “At the beginning of 2020, the entire makeup of how we were going to launch literally changed overnight,” says Keith Davidson, director of global over-the-top platforms at Sky. “But because we’d chosen the cloud and used AWS, we were able to pivot to remote work very quickly and still achieve our launch date.” 

Despite the challenges of the pandemic, Peacock’s soft launch in April 2020 went off without a hitch. Its full launch in July was similarly flawless, making an overwhelmingly positive first impression on users. “It’s strategically important for us to keep reliability and security at the forefront of our service architecture,” says Patrick Miceli, executive vice president and chief technology officer at Peacock. “As a result of our multiregion, distributed approach and the help of AWS, we were able to achieve Peacock’s technology goals rapidly and at scale.” 

Peacock’s cloud infrastructure on AWS has the flexibility to continue adapting for an even better user experience, as Davidson points out. “The ability to flex, adapt, and modify the infrastructure and service underneath is essential to us,” he says. “We are deploying literally every single day, and that will continue on for the life of the product.” In addition, Colin Innes, head of group reliability engineering at Sky, says the ease of building and testing infrastructure through AWS services makes fast and flawless implementation possible. “Our ability to automate and build things in the cloud—we couldn’t have achieved that velocity with what we were doing on premises,” he says. “It’s been a game changer for us.”

For Peacock’s technology team, the project launch was successful in part because of the help of AWS support. “We look at key resources like AWS as an extension of our team rather than a vendor with a contract. We have to be a team,” says Davidson. As for the effectiveness of AWS solutions, Black says Peacock’s success speaks for itself. “The proof is in the performance. We have a commitment to AWS because the system continues to deliver, and we and our users are pleased with the results.”