This is a senior engineering role with a genuinely novel context: the vast majority of feature code on this programme is generated by AI coding agents, not written by hand. As a Senior Full Stack Engineer, you are the final human quality gate in the delivery team – responsible for interpreting, reviewing, validating, and where necessary correcting agent-generated code before it is approved for merge and deployment.
AI Agent Output Review and Correction This is your primary responsibility on this programme:
- Review, validate, and correct AI-generated code across the full stack before approving for merge – assessing correctness, security, performance, and structural integrity
- Validate against the technical design – confirm agent output implements what was specified before approving for merge
- Classify agent output deviations – distinguish errors in agent execution from gaps in the specification, and communicate findings clearly
- Issue targeted correction instructions where output requires adjustment; directly patch agent output where a focused fix is more efficient
- Validate agent-generated tests for adequacy and correctness – identifying where coverage is insufficient or where tests pass trivially
- Escalate recurring failure patterns to the AI Engineer – providing the signal needed to improve agent contracts programme-wide
Full-Stack Feature Accountability
- Take end-to-end accountability for features – from Angular UI through Node.js Lambda API to DynamoDB/PostgreSQL data layer and AWS infrastructure
- Identify and resolve performance bottlenecks, security vulnerabilities, and structural issues in agent generated code before merge
- Maintain API specifications and contribute to architecture decision records for team-level decisions
- Integrate agent-generated outputs into CI/CD pipelines and ensure deployment readiness
- Collaborate with Systems Analysts to understand acceptance criteria and raise specification questions before agent execution begins
- Participate in sprint ceremonies – contributing a technical perspective on agent output quality
- Work with QA engineers on test data requirements and edge case identification.
