DishTV Launches an End-to-End OTT Platform on AWS

Cloud technology was a given for the new platform. Demand for OTT content is highly variable, with workloads spiking 10 times or more from day to day. Scalability offered on the cloud would have been harder to accomplish on premises, where a majority of DishTV’s workloads were hosted prior to this project.

The company already had some applications running on the Amazon Web Services (AWS) Cloud, and it chose AWS for this project based on price, performance, and its breadth and depth of service offerings. “The breadth of AWS solutions is much more robust than their competitors’ offer. Other companies are good in specific things but cannot give us what we need for multiple use cases for our applications,” Tyagi says. DishTV uses AWS services for its end-to-end OTT solution, from ingestion of content and processing to advanced use cases that leverage machine learning through to delivery.

To affordably house massive datasets on the platform, DishTV takes advantage of Amazon Simple Storage Service (Amazon S3). For video processing, it uses AWS Elemental, and for content delivery, it relies on Amazon CloudFront. The team also deployed AWS Lambda to run the startup code on the Watcho application’s backend.

When designing the OTT infrastructure, the team noticed that their Amazon Elastic Compute Cloud (Amazon EC2) servers—the initial choice for their base architecture—were often sitting idle. They decided to deploy Docker containers to improve resource utilization and chose Amazon Elastic Container Service (Amazon ECS) to run and scale containerized applications. Containers have cut costs by firing up servers only when necessary, using AWS Auto Scaling and Elastic Load Balancing to expand resources during big entertainment events such as national cricket matches. “That’s a huge advantage in terms of reducing our overall cost of ownership for the project,” Tyagi says. Containers also make it easier to isolate target environments when small changes are required, without impacting the entire infrastructure.

DishTV launched Watcho in January 2019, just seven months after development had started. “Building an end-to-end solution on AWS helped speed up time-to-market,” says Tyagi. “Typical OTT development time for operators our size, also hosted on public or private clouds using a range of providers, range from 12 to 14 months.” Several members of DishTV’s team as well as those of its implementation partner, 99Array, completed AWS certifications prior to starting the project, which eased the configuration process and ensured they were following best practices from the start. The fact that many supply chain partners, such as the company handling packaging, were already integrated with AWS also sped up the process by at least a month.

Since going live, Watcho has attracted an average of 7,000 new users daily, with significant organic traffic from existing DishTV subscribers. “The OTT platform allows us to reach a larger base of consumers, cater to them in a digital format, explore avenues to expand our revenue stream, and reduce churn while increasing the stickiness of our overall brand,” Tyagi says. “Being on the AWS Cloud helps us create, develop, and deploy software at a much faster pace compared to what we have been doing on premises and achieve larger scale in terms of volumes of customers as well as content distribution, which we weren’t able to do earlier.”

DishTV has been able to create new revenue streams because of the increased advertising dollars flowing into the platform, and it aims to bring Watcho’s original content to Indian diasporas through various partnerships across the globe soon. DishTV also plans to launch new services on Watcho such as a user-generated content module called Watcho Community, which aims to empower content creators to showcase their content and engage with their followers on the platform.

Although the company is still in the early stages of data collection, it is already gathering useful quantitative insights such as the days and times customers view OTT content. Such data helps DishTV marketers understand how and when to communicate with customers and upsell their products. The data also aids in market segmentation and mapping content preferences. Ultimately, the company’s goal is to streamline its content offering to match demand. Content production is a major cost center and tailoring its programming to customers’ tastes—cutting programs with fewer viewers—could save DishTV thousands of dollars per month.

The team is also experimenting with artificial intelligence (AI) and machine learning (ML) to fuel innovation. “AWS offers a strategic advantage from a digital transformation perspective, using new services such as Amazon Lex, Amazon Polly, and Amazon Rekognition, all of which would be too cumbersome to create and manage on premises,” Tyagi says. DishTV uses Amazon Lex to enable mobile voice commands and Amazon Rekognition to flag abusive language on its user-generated content feature. It is beta testing Amazon Polly and Amazon Transcribe for subtitling speech to text and vice versa, and it’s exploring the use of Amazon Personalize to build a recommendations engine.

Tyagi concludes, “Overall, AWS has a very proactive approach towards customers, which I really appreciate. They keep coming up with new types of engagements to help us understand the gaps in our technology and how best to optimize.”