Spot
94 Case Studies
A Spot Case Study
Duolingo, the Pittsburgh-based language-learning platform with over 300 million users, faced rapidly rising AWS ECS costs after moving to microservices. They struggled with underutilized reserved instances, inability to mix instance types/sizes for better packing, and needed a solution that would work with their Terraform-based deployment workflow. To address this, Duolingo engaged Spot (using Elastigroup).
Spot’s Elastigroup automatically prioritized reserved instance usage, leveraged Spot market scoring to select the most cost-effective instance types across AZs, and used the Elastigroup ECS Autoscaler (with Tetris Scaling and managed headroom) to scale reliably. By integrating Elastigroup with their Terraform pipeline, Duolingo cut AWS compute costs by 65%, simplified operations, and avoided major workflow changes.