Duolingo English Test - AWS Architecture

aws · architecture diagram.

About This Architecture

Duolingo English Test leverages AWS to deliver a secure, AI-powered language assessment platform with isolated infrastructure and real-time scoring. Test takers connect through AWS WAF and CloudFront for DDoS protection and global content delivery, routing to an ECS/EKS test delivery engine that orchestrates session state in DynamoDB and secure recording storage in S3. Amazon SageMaker performs real-time scoring and anomaly detection, while Amazon Polly generates neural audio stimuli for listening comprehension sections. AWS Organizations and IAM enforce compliance boundaries and least-privilege access across the isolated DET account, ensuring test integrity and regulatory adherence. Fork this diagram to customize security policies, scaling parameters, or ML model configurations for your own high-stakes assessment workload. This architecture demonstrates how to combine containerized delivery, managed ML services, and strict isolation controls for mission-critical testing platforms.

People also ask

How does Duolingo English Test use AWS services to deliver secure, scalable language assessments with real-time ML scoring?

The DET architecture uses AWS WAF and CloudFront for security and global delivery, ECS/EKS for containerized test delivery, SageMaker for real-time scoring and anomaly detection, DynamoDB for session state, and S3 for secure recordings. AWS Organizations and IAM enforce compliance boundaries and least-privilege access across an isolated account.

Duolingo English Test - AWS Architecture

AWSadvancedECSEKSSageMakerassessment-platformsecurity-architecture
Domain: Cloud AwsAudience: AWS solutions architects designing secure, scalable assessment platforms
1 views0 favoritesPublic

Created by

March 8, 2026

Updated

March 25, 2026 at 1:21 AM

Type

architecture

Need a custom architecture diagram?

Describe your architecture in plain English and get a production-ready Draw.io diagram in seconds. Works for AWS, Azure, GCP, Kubernetes, and more.

Generate with AI