“With hyper-growth that we were (and are seeing) on a year-on-year basis, we needed a highly scaled solution that can be elastic as per our traffic patterns and data volumes. We were also looking to get reliable out of the box software/infrastructure as a service so that we could focus on our core product. Cloud Technologies fitted the bill perfectly,” said Abhishek Ravi, CIO, Dream Sports.
According to Ravi, some key aspects that are considered for going with the cloud are scalability, elasticity, performance, reliability, resilience, security and cost.
Most of Dream Sports’ application workloads are on the cloud. Dream11 runs mainly on Amazon Web Services. However, the company is also using (and exploring) other cloud services for other Dream Sports companies and use cases.
“Cloud has really helped us to plan, develop and scale our product without worrying about the performance, availability and cost of ownership. We could quickly test out our features, scale tests in load/stress environments and ship them to our users to give them the best experience. With managed services available on the cloud, our teams could and continue to focus on core products and thus, derive maximum efficiency,” Ravi added.
The company intends to remain cloud-native in the future too. “With our data volumes increasing day by day and newer solutions evolving in cloud space, we want to use the cloud at the optimum.”
The company’s strategy is multi-cloud as it selects the right solution for the different use cases as per the requirements.
Ravi believes that a multi-cloud strategy should be well thought through so that the best cloud technology for the right use case can be provisioned. It also helps to optimise the infrastructure and thus, the cost.
“For us, our multi-cloud strategy pushes us to do “Cloud Agnostic” architecture which is very important to reduce unwanted expenses, creating two centres of excellence, etc. We are making a conscious effort towards this. It is helping us to architect the best right from the beginning, depending on our scale and requirements. Hence, we are seeing good optimisation on all fronts,” he said.
In the coming months, Dream Sports aims to upscale its tech advancements and expand tech infrastructure.
“We leverage Big Data, Analytics, Artificial Intelligence and Machine Learning to focus on every aspect that makes sports better. We are heavily experimenting with push architecture to serve information to the users in real-time. We are also very advanced on containerization which has reduced our infrastructure requirements drastically,” Ravi maintained.
The company is also working on several tech initiatives such as a concurrency prediction model to predict hourly concurrency on the Dream11 platform, and a fraud detection system to identify & mitigate users creating multiple/ duplicate accounts on the platform to abuse referral or promotional cash bonus schemes.
To ensure a quality user experience during peak traffic, Dream Sports also stress tests every feature that is released for a smooth experience at scale. “We have a testing framework that simulates any kind of traffic load with real life-like patterns. This gives a high degree of assurance that the backend would behave exactly as expected,” Ravi added.