QA Test Analyst, Bluecrest Wellness
- Resolved a long-standing defect in blood pressure result handling, tracing the underlying system behaviour and implementing a fix that removed around 100 unnecessary customer reassessments a year.
- Design and maintain automated end-to-end testing with Cypress and Cucumber, managed in Git and run in Bitbucket CI/CD pipelines for business-critical services.
- Designed and built an AI-assisted QA agent that gathers ticket context and acceptance criteria, reviews code changes and their blast radius, checks deployment and regression status, and produces risk-based test plans. Uses reusable skills and specialist sub-agents, with benchmarked smaller models handling data gathering to cut token cost, and human review kept throughout.
- Prototyped AI-enabled decision tools, including a configurable decision tree for exploring how calculation changes affect the products shown to prospective customers, and a concept using customer data to generate tailored recommendations.
- Investigate business and production problems across frontend, backend, API and database layers, using SQL, REST APIs, logs and code analysis to find root causes and define fixes.
- Work with HL7 message data in integration testing and incident investigation, validating messages and tracing failures through error logs.
- Contribute directly to Angular, TypeScript, PHP and SQL code, taking work from investigation through implementation, testing and release.
- Delivered an internal Stripe subscription-management feature so Contact Centre agents can cancel subscriptions without opening Stripe. Spotted missing backend data mid-build, specified the extra API, worked with a backend developer on the endpoint and integrated it.
- Act as Accessibility Champion, supporting colleagues with accessible design and development and building an interactive accessible staff training course.
- Act as AI Champion, supporting responsible experimentation with AI tooling and its practical adoption across the team.