Spacemind

Tóm tắt công việc

Own the quality assurance function end-to-end for a high-growth enterprise engineering organisation on a 500M€ growth mandate. You are not a gatekeeper at the end of a pipeline, you are embedded from requirements to release, shifting quality left and automating everything that can be automated.

You write code. You build frameworks. Manual testing is your documented exception, not your default.

We are building AI-native. You will validate traditional software behaviour and AI outputs: LLM responses, agent workflows, inference accuracy, and failure modes unique to non-deterministic systems. Nothing ships without your sign-off. You report directly to the Tech Lead and CEO.

SCOPE & IMPACT

QA Strategy

• Own org-wide QA strategy: coverage targets, risk frameworks, release gates, and escalations
• Define and enforce a Definition of Done - no sprint closes without quality criteria met
• Own bug triage: severity SLAs, P0/P1 post-mortems, defect escape tracking, metrics published continuously
• Build and own end-to-end test automation: E2E (Playwright/Cypress), API (Pact/Newman), Performance (k6/Gatling), Security (OWASP ZAP)
• All automation integrated into CI/CD, no merge to main without gate passing; 80% automation ratio within 6 months
• Design test strategies for non-deterministic AI systems: accuracy thresholds, hallucination detection, failure taxonomies
• Build evaluation harnesses for LLM responses and agent workflows end-to-end
• Create regression suites that detect silent model degradation on update
• Collaborate with AI engineering on production observability hooks for quality signals

Release Governance & Security

• Own the release sign-off process, no feature goes to production without documented QA approval
• Define environment promotion criteria (dev → staging → production) with automated quality gates at each transition
• OWASP Top 10 coverage; SAST on every PR; DAST on every release candidate; GDPR/PII flow validation
• Full test traceability and documented sign-off for the clinical verification application (regulated domain)

Shift-Left & Developer Enablement

• Embed QA from requirements stage: review specs, challenge acceptance criteria, flag testability gaps before code is written
• Pair with developers on test-driven approaches; run QA onboarding for every new engineer

The Quality Bar You Will Set & Hold

• Zero critical or high defects to production - every escape triggers a post-mortem
• ≥80% automation ratio within 6 months - manual testing requires documented justification
• 100% critical path API contract coverage - breaks caught in CI, not staging
• SLOs defined and enforced per product track - regressions detected before users feel them
• AI output accuracy thresholds defined per use case - every AI feature has a test harness
• Every release signed off, documented, and traceable. SAST on every PR. DAST on every release candidate

• Competitive salary Full health & social insurance (retirement, sick leave, maternity) 
• Paid annual leave + guaranteed Tet bonus Gym membership, sports activities, team events 
• If you’re driven, proactive and curious about smart, AI-enabled ways to scale e-commerce, join us. 
• Global exposure, real responsibility, tangible results.

Function Ownership

• Owned a QA function end-to-end: strategy, standards, tooling, and accountability - not just executed within one
• Demonstrated ability to hold engineering teams to quality standards, not just report on them
• Shipped quality outcomes in a product organisation (not agency or outsourcing)

Automation Engineering > Non-Negotiable

• Writes test automation code daily. Builds frameworks. Does not manage manual testers.
• E2E: Playwright or Cypress | API: Pact or REST Assured | Performance: k6 or Gatling | Security: OWASP ZAP
• CI/CD: writes and modifies pipeline configurations (GitHub Actions or GitLab CI)
• Scripting in at least two of: TypeScript, Python, Java - for framework authoring
• Domain depth: PostgreSQL, REST APIs, OAuth2/OIDC, GDPR/PII data flow validation

AI-Native Requirements

• Uses AI tooling (Cursor or equivalent) as primary development environment for test code
• Has designed and executed test strategies for non-deterministic AI systems - shipped, not theoretical
• Applies AI to accelerate coverage analysis, edge case generation, and defect pattern recognition
• Stays current with the rapidly evolving field of AI system testing methodology

Communication

• Excellent written and spoken English - crisp bug reports, clear standards, unambiguous escalations
• Able to present quality status and risk to C-level stakeholders without technical translation

Nice-to-Have

• Regulated or clinical software environments (IEC 62304 or equivalent)
• Visual regression testing (Percy, Chromatic); chaos engineering; multi-country compliance testing