ELSA Scales Quickly and Reliably on the AWS Cloud to Support Millions of Users

ELSA Corp. (ELSA) uses speech technology and artificial intelligence (AI) to help non-native English learners improve their speech and pronunciation via short, app-based lessons. Its ELSA Speak app is currently localized in nine languages and accessed by users across over 100 countries for learning English. Vietnam served as its first test market in 2016.

Today, ELSA’s user base numbers 15 million learners, part of the growth driven by demand for online learning from pandemic community lockdowns globally.

To help process large amounts of data from the high volume of daily user activity, the company needed to ensure its cloud-based computing platform continues to allow for flexible compute, database, and storage configurations. In addition, ELSA’s researchers and developers required easy access to real-time and stored user-generated data to continually improve its proprietary speech algorithms.

According to Xavier Anguera, co-founder and chief technology officer of ELSA, his company uses the Amazon Web Services (AWS) Cloud as its global computing infrastructure. This has helped it deliver content quickly to its users, and process over four million user recordings daily.

In September 2019, ELSA worked with AWS to containerize its workloads with Amazon Elastic Kubernetes Service (Amazon EKS) on a mix of Amazon Elastic Cloud Compute (Amazon EC2) On-Demand and Spot Instances. This equipped ELSA to provision memory and compute resources for diverse development work, and optimize workload cost with performance.

To ensure low latency and high availability of its app, ELSA adopted a combination of Amazon DynamoDB and Amazon Relational Database Service (Amazon RDS) as its core databases to support a rapid user growth, while content are streamed to its users via Amazon CloudFront.

The company uses Amazon Simple Storage Service (Amazon S3) to store almost 200TB of content, including two million hours of user-generated audio recordings. Amazon Athena, a serverless query service, is deployed to run queries and extract insights easily and directly with data stored in Amazon S3.  

ELSA has replicated its cloud infrastructure across three AWS Regions to cater to its growing user base and locations.

In teaming with AWS, ELSA can focus on research, business strategy, new market expansion, and customer satisfaction.

“The AWS Cloud ensures low latency, and fast query and extraction of real-time data,” shares Anguera. “This helps us maintain high levels of user satisfaction since our launch. Today, our ELSA Speak app has achieved 4.6 stars and 4.7 stars ratings respectively in the Google Play Store and Apple Store, from a total of 300,000 user reviews.”

To learn more, visit aws.amazon.com/education/.