Alt/S Case Study

Alt/S is a tech startup based in New York and Silicon Valley that specializes in data analytics and insights, moving the focus from reporting to actionable insights, predictive analytics, and next-gen applications. Alt/S’s Sportify platform is a sports-analytics service that uses modern data collection and modeling techniques to develop insights-driven analyses of the world’s largest sports. Sportify’s insights and content—powered by its innovative machine-learning-based platform—have reached more than 20 million fans since launch in June 2016.

• Alt/S wanted to build its Sportify sports-intelligence platform to support real-time analytics, high scalability, faster development of products and features along with lower costs.
• Sportify required large-scale analytics computation and pattern mining, rapid hypothesis validation, modelling game states and building team and player performance reports and predictions.
• Alt/S wanted to develop highly-scalable APIs and backend capabilities to support different applications, from video to fantasy sports to broader big-data analytics.
• Needed to seamlessly consume data from a variety of data sources with varying access protocols and structures.
• Hoped to maximize time spent on innovations and core product development and eliminate time spent on infrastructure provisioning and maintenance tasks.
  

• The complete backend platform was built using an event-driven, completely serverless microservices architecture using AWS Lambda and related Amazon Web Services (AWS) products.
• Different microservices were created using AWS Lambda to perform operations, including data acquisition, transformation, ingestion, analytics, a number of machine-learning activities such as pattern mining, supervised and unsupervised learning, predictions, and visualization.
• RESTful API were developed using Amazon API Gateway and AWS Lambda.
• Amazon Cognito authenticates users to various backend services.
• Various data sources are supported including Amazon Relational Database Service (Amazon RDS), Amazon Redshift, Amazon EMR, Amazon Simple Storage Service (Amazon S3), various third-party APIs, and public-domain Wikipedia information.
• Microservices automatically scale based on load and triggers from different event sources, including Amazon S3, Amazon Simple Notification Service (Amazon SNS), and Amazon API Gateway.
• Front-end web applications are deployed and managed using AWS Elastic Beanstalk, which provides simplified management and automatic scaling.
• Amazon CloudFront is used for content delivery.
  

• Zero time is now spent deploying or maintaining server instances, as Sportify’s entire backend platform is based on AWS Lambda.
• The solution scales automatically.
• New products and features are delivered faster since each distinct feature can be developed as a separate microservice that can be tested and deployed independently. A new product comprising a web application, secure RESTful API, user management, data store, predictive analytics and backend workflows was built, tested, deployed and scaled in less than four weeks. This was primarily made possible by leveraging AWS Lambda and its integration with multiple AWS services.
• Services run only on demand based on usage and don’t need to run throughout the day, which lowers costs.
• Developer productivity has improved as their time is primarily spent developing core product innovations.