Quality assurance (QA) automation has long been a bottleneck in software delivery, plagued by brittle test frameworks that demand specialized coding skills and constant maintenance. Amazon Nova Act, a new service from AWS, aims to eliminate these inefficiencies by introducing agentic QA automation—a system that mimics human interaction to navigate applications through natural language and visual understanding rather than brittle code-based selectors. This approach reduces test maintenance overhead while accelerating software delivery cycles by aligning testing directly with business requirements.
The Challenge of Traditional QA Automation
Traditional QA automation frameworks rely on UI selectors, element identifiers, and structural references to interact with applications. When developers refactor code or designers adjust layouts, tests frequently break—even when functionality remains intact. This maintenance burden stems from a disconnect between how teams work: product managers define requirements in business terms, developers implement features, and automation engineers write code-based tests. The result? Teams spend more time fixing broken tests than delivering new features.
Amazon Nova Act addresses this gap with a custom computer use model that interacts with applications the same way users do—through natural language and visual understanding. By removing code-dependent selectors, it reduces test maintenance and democratizes test management, enabling teams to focus on delivering features rather than debugging infrastructure.
How QA Studio Leverages Amazon Nova Act
AWS’s QA Studio serves as a reference solution built on Amazon Nova Act, offering a web frontend, API, and CLI for managing QA automation. The platform uses serverless AWS infrastructure to execute tests reliably at scale, with capabilities including:
- Natural language test management: Teams define tests in the same language used for product requirements, eliminating the need for code-based test scripts.
- Visual navigation: Tests adapt automatically to UI changes (e.g., button relocations or layout adjustments) by relying on visual context rather than brittle selectors.
- End-to-end test visibility: Trajectory logs provide step-by-step visual reasoning, screenshots, and natural language explanations of test execution—simplifying debugging.
Technical Architecture Behind QA Studio
QA Studio’s serverless architecture integrates multiple AWS services to deliver scalable, cost-efficient test automation:
- Amazon Nova Act: Agentic UI automation engine
- Amazon Bedrock: Foundational models for test case generation and AgentCore Browser
- AWS Lambda and Amazon DynamoDB: Backend logic and test data storage
- Amazon ECS with Fargate: Containerized test execution
- Amazon SQS and EventBridge: Reliable scheduling and queue-based processing
- Amazon CloudFront and S3: Frontend delivery and artifact storage
This design ensures automatic scaling and pay-per-use economics, making it accessible for teams of all sizes.
Key Benefits of Agentic QA Automation
By shifting from code-centric to agentic automation, teams gain:
- Reduced maintenance: Tests adapt to UI changes without manual updates.
- Faster delivery: Teams focus on features, not test infrastructure.
- Improved collaboration: Natural language test definitions align with business requirements.
- Enhanced debugging: Visual trajectory logs provide clear insights into test behavior.
Read more: aws.amazon.com