For omnichannel game provider Playson, delivering games at low latencies is key to keeping players engaged and maximizing their betting opportunities. Since its founding, the company has quickly grown, expanding to new customer bases while regularly releasing fresh titles. To keep up with demand, the Playson team needed to scale its gaming service, which was difficult with its monolithic, multicloud architecture. To enhance its agility, Playson decided to go all in with one cloud service provider and chose Amazon Web Services (AWS) for its scalability and reliability. In 8 months, the Playson team migrated its workloads from 2,000 servers to AWS. As a result, Playson can deliver a faster, more reliable gaming experience to players while simplifying its monitoring and governance capabilities. In the next year, Playson fully rebuilt its gaming service, xPlatform, with a microservices architecture using Amazon Elastic Kubernetes Services (Amazon EKS), a managed Kubernetes service that runs Kubernetes on AWS and on-premises data centers. By taking a microservices approach, the company has reduced spin time and improved scalability during prime gaming times, which has led to an increase in bets of up to 200 percent per second. Casino Gaming Software Opportunity | Enrolling in AWS MAP to Modernize a Monolithic Architecture for Playson Playson offers a portfolio of 69 cutting-edge, high-quality casino games, including titles like Royal Coins 2 and Lion Gems. The company provides its games to thousands of iGaming businesses, which can connect with Playson’s turnkey gaming service through direct integrations. Playson has seen these collaborations grow by 8 percent year over year since 2015. With its monolithic architecture, Playson would set up one server for each client it acquired. As the company grew, it became difficult to sustain this infrastructure model. Further, the company experienced downtime each time it released a title, which impacted the player experience. “Most players compare their online experience with playing slot machines in a casino,” says Oleksii Mylotskyi, chief technology officer at Playson. “So, everything that we offer has to be blazingly fast.” In addition to its on-premises environment, Playson had a portion of its workloads running on services from multiple cloud providers, including AWS. As the company continued to grow, it wanted to achieve high availability and scalability by decoupling its applications from each other. “We wanted to transition to a microservices architecture and extensively use AWS services, with the intention of improving our system’s resilience, agility, and scalability,” says Mylotskyi. “Our desire was to break up our applications into smaller, autonomous services to facilitate a more efficient development cycle, provide superior fault isolation, and prompt feature releases.” Playson understood the strategic advantage that fully committing to AWS could offer to its service. The comprehensive suite of AWS services, coupled with robust and scalable infrastructure, presented an opportunity to significantly enhance the value that Playson could provide to its customers. To plan for a zero-downtime migration, the company enrolled in the AWS Migration Acceleration Program (AWS MAP), which helps accelerate businesses’ migration and modernization journeys with an outcome-driven methodology. kr_quotemark Using AWS, we can build a more stable service while providing the best experiences for players who are located in different regions.” Oleksii Mylotskyi, Chief Technology Officer, Playson Solution | Reducing Spin Time and Increasing Betting by 200% with a Microservices Architecture on AWS The Playson team learned about AWS best practices through AWS MAP, helping it design an architecture that lent itself to low-latency requirements with input from subject matter experts. “The AWS account managers and solutions architects oversee certain domains and customer bases,” says Oleksandr Rudenko, platform tribe leader at Playson. “They know how the gaming industry works and have given us good advice.” To avoid downtime, the Playson team and AWS team worked together to migrate each of the company’s 2,000 servers to the cloud one by one. During the migration, Playson adopted Amazon CloudFront, a content delivery network service built for high performance, security, and developer convenience, to deliver its games at scale, and now processes incoming API requests using Amazon EKS. For storing its key-value data, the company uses Amazon MemoryDB for Redis (Amazon MemoryDB), a Redis-compatible, durable, in-memory database service for ultrafast performance. As an object database, Playson has adopted Amazon DynamoDB, a fast, flexible NoSQL database service for single-digit millisecond performance at any scale. By building a microservices architecture on AWS, Playson has reduced its games’ spin time by up to 350 milliseconds, which is in the 95th percentile. This reduction has meant a smoother, more immersive player experience. “We offer the opportunity for players to play our games faster now,” says Rudenko. “We even added a turbospin feature because we have the technical capability.” Because players can partake in more rounds over a shorter period, they tend to bet more often, too. In some cases, Playson’s clients saw the number of bets per second increase up to four times, increasing their revenue streams and overall profitability. Another benefit of its microservices architecture is that Playson no longer experiences downtime when launching new games, speeding up its time to market. “We can deliver games globally up to 10 times faster than before,” says Mylotskyi. Outcome | Enhancing Resiliency and Availability Using Multiple AWS Regions Playson will continue expanding to new customer bases and developing games, having architected a scalable, agile gaming service. Because availability is mission critical in the gaming industry, Playson is working toward deploying its infrastructure across multiple AWS Regions. “Using AWS, we can build a more stable service while providing the best experiences for players who are located in different regions,” says Mylotskyi. About Playson Omnichannel casino game developer Playson has released 69 cutting-edge, high-quality gaming experiences, which the company delivers to thousands of iGaming businesses through its turnkey gaming service.

Overview Game development services company Upsoft set out to simplify game distribution for developers by offering PatchKit—a content delivery network (CDN), hosting service, and launcher that has everything set up for developers so that they can focus on their games. However, the company’s attempts to roll its own infrastructure for this solution proved unreliable and required too much work, constant maintenance, and high costs. It decided to migrate to Amazon Web Services (AWS) to take advantage of fully managed solutions and reliable infrastructure. Now, Upsoft is using Amazon CloudFront, which securely delivers content with low latency and high transfer speeds, as the foundation for its CDN and other PatchKit services. The company has improved content delivery speeds, enhanced reliability, and instilled trust in its customers. Gaming Keyboard Opportunity | Using Amazon CloudFront to Enhance Reliability and Instill Trust in Clients for Upsoft Based in Poland, Upsoft simplifies game hosting, content management, and more so that game developers can focus on the creative process. The company’s PatchKit solution includes PatchKit Launcher, which handles game activation and user authentication, and PatchKit CDN, which provides fast, secure content distribution with no additional configuration required. The challenge for Upsoft was building PatchKit on a reliable foundation that its customers would know they could trust. “With a service like ours, the biggest issue is trust,” says Piotr Korzuszek, chief executive officer (CEO) of Upsoft. “Our clients need to trust that we’ll stick around after they decide to use us.” Upsoft had built the initial versions of PatchKit by using multiple smaller CDNs and its own virtual private servers. However, this required a lot of maintenance and resulted in high distribution costs. Worse, these attempts didn’t prove reliable, and customers noticed. “The moment we published our first post about PatchKit on social media, our service went down,” says Korzuszek. Upsoft wasn’t happy with the downtime its clients were experiencing, partly as a result of electronic service providers blocking its virtual private servers and interrupting software downloads. Players depend on a service being available at all times, and any downtime creates a bad user experience. Upsoft needed a more reliable foundation for PatchKit. After testing several cloud providers, the company decided to migrate the solution to AWS because of its reliability and compatibility with its specific needs. Upsoft had already been running much of the rest of its infrastructure on AWS for years. As a result, the migration of its CDN to Amazon CloudFront took just 1 month. The company did extensive testing to make sure everything would work properly on the new system. At one point, it was running two CDNs—the old version on its custom infrastructure and the new version on Amazon CloudFront. It used a copy of all its resources on Amazon CloudFront to test each step until it was ready to migrate fully to the new system. PatchKit has been available since 2015, and the migration to Amazon CloudFront was complete in 2022. Immediately, the migration to AWS had a significant impact. “2022 was the first year we had profit from our service,” says Korzuszek. kr_quotemark This was game changing, because this was the moment when we could tell our clients: Hey, you can trust us because we’re hosting on AWS." Piotr Korzuszek CEO, Upsoft Solution | Accelerating Content Publishing by 50% Using AWS After migrating to AWS, Upsoft quickly saw significant improvements in its data transfers, availability, and the all-important factor: client trust. “This was game changing, because this was the moment when we could tell our clients, ‘Hey, you can trust us because we’re hosting on AWS,’” says Korzuszek. The biggest advantage of using Amazon CloudFront is that Upsoft doesn’t have to use different points of presence all over the world to deliver content globally. And it’s still delivering content faster than it used to. Average data speeds in Poland are about 100 Mbps. On Amazon CloudFront, PatchKit is delivering content at the full bandwidth speed available. Further, Upsoft has accelerated content publishing by more than 50 percent and improved the stability of this process. Upsoft’s clients experienced better reliability and fewer interruptions. “All of the incorrect content blocking stopped as soon as we switched to Amazon CloudFront,” says Korzuszek. The company estimates that, on its previous solution, about 10 percent of users complained about data transfers. After switching to Amazon CloudFront, the company estimates that less than 1 percent experience any issues. Upsoft also improved its authentication process for specific resources using AWS Lambda, which businesses can use to run code without thinking about servers or clusters. It’s handling JSON web tokens for authentication requests more reliably using Lambda@Edge, a feature of Amazon CloudFront that helps businesses run code closer to users. And to locate any code issues, Upsoft set up Sentry, an application that developers can use to monitor application code health. “So, if anything happens, we can just check out what happened, try to optimize it, and fix it,” says Korzuszek. Overall, the company has less infrastructure to manage and maintain. “We used to spend 25 percent of our time each month on software upgrades, bug fixes, and general server maintenance,” says Korzuszek. Now, Upsoft is using just two Amazon CloudFront distributions—Europe and North America—to serve most of its clients. “The clients are really happy about our data transfer speeds,” says Korzuszek. Upsoft communicated closely with the AWS account manager and solutions architect throughout the migration. “We had many online calls with the AWS team,” says Korzuszek. “I was really happy that I could speak to them whenever I needed, and I felt quite confident with this support.” And when a billing issue arose, AWS communicated with the company and quickly corrected the information. “I was very happy with the support,” says Korzuszek. “Everyone handled the case professionally.” Outcome | Delivering Fast Data Transfers and High Availability for Clients and Players Upsoft’s migration to Amazon CloudFront is complete, but the company looks forward to using AWS to improve more aspects of its infrastructure and offerings. “Our biggest advantage is that we’re offering a service that’s based on solid foundations, on a CDN that’s one of the most reliable in the world,” says Korzuszek. With the trust of its clients and the ability to provide high availability to players, Upsoft is positioned to continue to grow and improve PatchKit. “Everything went so exceptionally well that we want to continue using AWS and Amazon CloudFront,” says Korzuszek. “We wouldn’t go anywhere else.”

Overview As a global leader in travel, Skyscanner Ltd. (Skyscanner) made the strategic decision to operate in one cloud environment as a means to future-proof its environment and identify opportunities for cost savings. Because the company serves 100 million people each month through its travel marketplace, fault tolerance was a high priority for Skyscanner while consolidating its technological stack. Skyscanner had already migrated its front-facing applications from its data center to Amazon Web Services (AWS) in 2017. Based on its experience, the company wanted to standardize its content delivery network (CDN) on AWS. So, the Skyscanner team adopted Amazon CloudFront, which securely delivers dynamic and static content with low latency and high transfer speeds. The Skyscanner team also built a serverless image handler that compresses static content using Amazon CloudFront, helping the company achieve 50 percent cost savings across its total CDN usage. Airplane flying over business skyscrapers, high-rise buildings. Transport, transportation, travel. Sun light on blue sky. Opportunity | Using Amazon CloudFront to Optimize the Technology Stack for Skyscanner Skyscanner is a global leader in travel that connects over 100 million travelers each month with more than 1,200 trusted travel partners so that travelers can find the best flight, hotel, or car-hire options. Founded in 2003, Skyscanner has offices worldwide, in Europe, Asia-Pacific, and North America, where traveler-first innovations are developed and powered by data and insights. The company is committed to helping shape a more responsible future for travel in collaboration with its partners and by making use of the latest technology so that every traveler can explore the world effortlessly for generations to come. As Skyscanner has grown to serve over 110 market domains, the company wanted to support engineering efficiency and productivity while optimizing its cloud spend. Although Skyscanner had invested in AWS technologies, it used a fully managed CDN solution from another provider. “One of the major challenges of this project was that we were untangling almost a decade’s worth of root configurations that our team had not implemented,” says Stuart Ross, senior engineering manager at Skyscanner. Another challenge that the Skyscanner team faced was migrating its CDN to AWS without degrading the customer experience. On any given day, Skyscanner can receive up to 1.5 billion API requests, representing about 24 TB of data. With such high demand, it was essential to avoid global incidents and downtime. kr_quotemark The migration to Amazon CloudFront has simplified the management of our infrastructure footprint. There are far fewer moving parts, and it’s largely driven by AWS-managed services, which is great.” Stuart Ross Senior Engineering Manager, Skyscanner Solution | Configuring a Serverless Image Handler and Multiregion Deployment Using AWS CDK Skyscanner engaged the AWS team to create a proof of concept (POC) for Amazon CloudFront. “The AWS team was amazing,” says Andrew Aylett, senior software engineer at Skyscanner. “We had the opportunity to talk to subject-matter experts to determine which AWS services would be the best fit for our road map.” During the 3-month POC phase, the Skyscanner team built customized configurations, including a serverless image-management handler that automatically compresses static images in the most cost-effective format. “That aspect was previously managed by our CDN provider, and we wanted Amazon CloudFront to have the same capabilities,” says Rory McCann, senior software engineer at Skyscanner. To set up these configurations, Skyscanner used the AWS Cloud Development Kit (AWS CDK), giving its team the ability to define its cloud application resources using familiar programming languages. “AWS CDK was key to this project,” says Aylett. “Our teams could write code rather than writing infrastructure.” Skyscanner sourced code for its configurations from the AWS Solutions Library, which provides vetted solutions and guidance for business and technical use cases. By making these resources available to its engineering teams, Skyscanner configured Amazon CloudFront with 1,000 lines of code—a significant reduction from its previous solution, which had over 26,000 lines. Skyscanner also configured Amazon CloudFront for multiregion deployment, increasing its fault tolerance. “Our team can sleep at night knowing that if something happened, there would be another AWS Region where we could automatically direct our web traffic,” says Ross. Protecting its front-facing applications and website from distributed denial-of-service attacks was a priority, too, so the Skyscanner team implemented AWS Shield, a managed distributed denial-of-service protection service that safeguards applications running on AWS. The team activated AWS Shield Advanced so that it has near-real-time visibility into distributed denial-of-service events and 24/7 support from the AWS Shield Response Team. After completing the POC, the Skyscanner team migrated its front-facing applications and website to Amazon CloudFront in increments, starting with its less-trafficked market domains. “It built up our confidence to start pushing the rest of our traffic from our consumer-facing sites to Amazon CloudFront,” says Aylett. The migration took a total of 3 months to complete, during which the Skyscanner team experienced zero global downtime. Since then, the Skyscanner team has been able to scale its serverless image handler to three billion monthly API requests while maintaining an average cache-hit rate of 99.99 percent. And by running its image handler on a serverless architecture, the Skyscanner team reduced its CDN costs by 50 percent. Outcome | Future-Proofing Its Architecture for Blue-Green Deployments To continue innovating, the Skyscanner team plans to adopt a blue-green deployment strategy, which will help its team reduce deployment risk and quickly roll back changes by creating two identical independent environments for routing web traffic. The Skyscanner team can accelerate its efforts toward this goal with a streamlined, standardized stack on AWS. “The migration to Amazon CloudFront has simplified the management of our infrastructure footprint,” says Ross. “There are far fewer moving parts, and it’s largely driven by AWS-managed services, which is great.”