Automation QA Engineer

Description

Our client operates a SaaS platform that supports the daily work of railway station staff in Japan. The platform is built as a Super App that hosts separate Mini-Apps, and it has recently been extended with a common AI platform.

We are building a proof of concept of an Education Mini-App on this platform. Station staff use it on an iPhone to study operational manuals and answer quiz questions in short sessions during the working day. Questions and explanations are generated by the AI service from the manuals, approved by a person, and served to users as versioned content.

GlobalLogic delivers a mobile-first staff web UI, an admin panel and a thin backend-for-frontend (BFF) service that connects them to the AI service and stores question versions and learning history. The staff web UI is converted into a Mini-App with the Eki-Suite Mini-App SDK and runs inside the client’s native Super App on iPhone. The native app and its App Store release are outside the GlobalLogic scope.

Technology: React 19 and TypeScript for the staff web UI and the admin panel, .NET 10 and ASP.NET Core for the BFF, PostgreSQL 17, AWS, Playwright for end-to-end tests.

The work runs in four phases: discovery, walking skeleton, MVP implementation, and validation with real station staff. You will be the only QA engineer in a small cross-functional team and will cover both manual and automated testing.

#LI-MP2
#LI-Remote

Requirements

• 3+ years of QA experience with web applications, including at least 1 year of hands-on test automation
• Hands-on manual testing of mobile-first or responsive web applications 
• Confident use of browser developer tools, including Web Inspector: network, console, storage
• Test automation with Playwright in TypeScript: UI end-to-end tests and API tests
• API testing from an OpenAPI description with Scalar, Postman or a similar tool
• SQL for data verification, PostgreSQL preferred
• Experience testing against a mocked external service and configuring mock responses, WireMock or similar
• Ability to run a system locally with Docker Compose
• Experience running automated tests in a CI pipeline and working with Git
• Ability to plan and run testing independently without a QA lead on the team: test cases, test data, defect tracking, status reporting
• Ability to work with requirements that are still being defined during discovery
• Testing of products that use LLM-generated content
• Clear written defect reports and test documentation in English, Upper-Intermediate or higher

Job responsibilities

• Prepare the test approach and test cases for the PoC, decide what is tested manually and what is automated, and agree acceptance criteria for each user journey with the business analyst and the delivery lead
• Manually test the Self-Study and Quiz journeys of the staff web UI in browser 
• Test sign-in of PoC test users through AWS Cognito and the handling of missing, invalid and expired tokens by the BFF
• Test the BFF API against its OpenAPI description: request validation, storage of answers and progress, error mapping
• Configure WireMock.NET to imitate the AI service, including slow responses and failures, and check timeouts and retries in the BFF
• Repeat the integration tests against the real AI service when it becomes available
• Verify content integrity in the UI and in PostgreSQL: only approved question versions are served, scoring matches the approved answer, each attempt stores the question version it was answered against
• Build and maintain Playwright end-to-end tests for the main journeys of the staff web UI on the WebKit engine and for the admin panel
• Build and maintain automated API tests for the BFF with Playwright 
• Run the automated tests in the CI pipeline together with the DevOps engineer
• Check the Japanese UI for text rendering, truncation and line breaking
• Analyse defects using application logs in AWS CloudWatch
• Support validation sessions with station staff: prepare test data, record defects and observations
• Report manual test status, automated test results and open defects at the weekly demo