Quality Guild Charter
Quality Guild Charter
What is a Guild
A guild is a cross-team community of practice — not a traditional delivery team. Unlike a team, a guild has no dedicated headcount, no shared backlog, and no single manager. Its members belong to their respective engineering teams and continue to report through their existing management chains.
The guild exists to solve a coordination problem: with tens of autonomous teams, quality practices diverge unless there is a deliberate mechanism for alignment. The guild provides that mechanism — shared standards, regular knowledge exchange, and collective ownership of quality culture — without creating a centralized bottleneck.
How a guild differs from a team:
| Dimension | Team | Guild |
|---|---|---|
| Reporting | Members report to one manager | Members report to their own team's manager |
| Allocation | Full-time dedication | Part-time participation alongside team work |
| Backlog | Owns a delivery backlog | Owns standards and governance, not delivery |
| Headcount | Has dedicated headcount | No dedicated headcount — draws from existing teams |
| Hiring | Hires into the team | Does not hire — hiring is done by the teams that embed QEs |
| Output | Ships features or infrastructure | Ships standards, playbooks, and coordination |
Mission
Drive consistent quality practices across Duetto engineering through shared standards, developer coaching, and cross-team coordination — ensuring quality is a shared responsibility owned by every engineer, not a gate owned by a separate team.
Team Metrics
Primary Metric
| Attribute | Value |
|---|---|
| Metric | Escaped Defect Rate |
| Definition | Number of bugs that reach production per release, across all teams in the guild's scope |
| Baseline | TBD (establish in Phase 1 audit) |
| Target | 30% reduction from baseline within 12 months |
| Measurement | Jira defect tickets tagged escaped-defect + PagerDuty incident correlation |
| Cadence | Monthly review, quarterly deep-dive |
Secondary Metrics
| Metric | Baseline | Target | Measurement |
|---|---|---|---|
| DORA Change Failure Rate | TBD | <5% across all teams | GitHub Actions + incident tracking |
| Quality Gate Compliance | 0% (no gates enforced) | 80% of repos at Phase 2+ gates | CI pipeline audit (GitHub Actions) |
| Flaky Test Rate | TBD (estimated >5%) | <1% of total test suite | DataDog test visibility dashboard |
| Cross-Team Testing Standard Adoption | 0% | 100% of teams follow guild standards document | Quarterly self-assessment + CI enforcement |
Counter-Metrics (Guard Rails)
| Metric | Acceptable Range | Alert If |
|---|---|---|
| Developer Satisfaction with QE Process | >7/10 NPS | <6/10 in quarterly survey |
| CI Pipeline Duration (P90) | <15 min for PR pipelines | >20 min average over 1 week |
| Guild Meeting Attendance | >60% of members per session | <40% for 2 consecutive sessions |
| PR Merge Lead Time | No increase from pre-guild baseline | >25% increase sustained over 1 month |
Scope
In Scope
- Testing standards ownership — author, maintain, and evolve the shared testing standards document (unit, integration, E2E, contract, data quality, pipeline — definitions and expectations)
- Quality gate governance — define what blocks a merge at each phase (Foundation → Contracts → Performance → Optimization) and when to promote gates from warning to blocking
- Guild cadence — bi-weekly all-hands (both tracks), monthly track-specific deep dives (App/Platform + Intelligence), quarterly standards review
- Developer coaching model — define what "QE coaches developers" means in practice: test strategy reviews, pairing sessions, exploratory testing guidance
- Cross-team consistency — shared test naming conventions, flaky test policy and SLAs, pre-commit hook standards, coverage reporting expectations
- Two-track coordination — ensure App/Platform and Intelligence tracks share governance and cross-pollinate learnings while maintaining domain-appropriate practices
- Career growth — QE career ladder (L1-L7), role definitions, hiring profile guidance, onboarding for new QEs and engineers
- Quality culture — quality onboarding for new developers, quality metrics visibility, recognition of testing excellence
- AI code quality strategy — standards for QA of AI-generated code (50-70% of Duetto code), mutation testing adoption, AI-generated test review guidelines
Out of Scope
- Building test infrastructure — owned by the Quality Engineering Team (see TC-007)
- Writing product tests — owned by engineering teams; QEs coach, they don't replace
- Model accuracy validation — owned by ML engineers; Intelligence QE ensures validation infrastructure exists
- Incident response — owned by on-call engineering teams; guild tracks escaped defects as a lagging indicator
- Tool procurement and vendor management — guild recommends; Engineering leadership and Procurement approve
- Security scanning policy — owned by Security team; guild integrates security gates into CI pipelines
Active Initiatives
No active initiatives yet — to be defined once the guild is established.
Guild Members
Guild members are not assigned to the guild — they belong to their respective engineering teams and participate in the guild as part of their cross-team responsibilities. The table below describes the guild roles, not team assignments. Each person's primary reporting line remains with their team's Engineering Manager.
| Guild Role | Participation | Responsibilities within the Guild |
|---|---|---|
| Guild Lead | Part-time (EM or Senior QE) | Coordinates standards, career ladder guidance, meeting facilitation, cross-track alignment |
| Staff/Lead QE | Part-time (also leads Quality Engineering Team — see TC-007) | Technical strategy for guild standards, quality gate design, tooling decisions |
| App/Platform Embedded QEs | Part-time (3-5 people, hired into and reporting to their product teams) | Contribute domain-specific testing knowledge, share patterns across teams, co-author standards |
| Intelligence Embedded QE | Part-time (1-2 people, Phase 2, hired into and reporting to Intelligence teams) | Contribute data quality and ML testing knowledge, co-author Intelligence track standards |
| Quality Engineering Team | Part-time (2-3 people — see TC-007) | Report on infrastructure and tooling progress, gather requirements from guild members |
| Interested Engineers | Voluntary, open membership | Participate in guild meetings, contribute to standards, champion quality in their teams |
Hiring Guidance
The guild does not hire. Hiring is done by the teams that need QE capability — the guild defines the role profiles, career ladder, and interview criteria to ensure consistency, but headcount belongs to the teams.
| Phase | Team Hiring | Guild Support |
|---|---|---|
| Phase 1 (Months 1-3) | Product teams hire 2-3 App/Platform QEs; Quality Engineering Team hires Staff/Lead QE | Guild defines QE role profile, interview rubric, and onboarding plan |
| Phase 2 (Months 3-6) | Intelligence teams hire 1 QE; Quality Engineering Team hires 1-2 engineers | Guild defines Intelligence QE profile (Python + ML + testing blend) |
| Phase 3 (Months 6-12) | Product teams scale App/Platform QEs to 5 | Guild reviews hiring criteria based on Phase 1-2 learnings |
Stakeholders
- Engineering Leadership (VP/Directors): Approve guild formation, quality gate enforcement levels, and QE headcount allocation to teams
- Engineering Managers: Embed QEs in their teams, support coaching model, attend monthly quality metrics reviews
- Tech Leads: Partner with QEs on test strategy, champion developer testing responsibilities
- Product Management: Align quality targets with product goals, understand trade-offs between speed and quality gates
- Intelligence Team Leads: Partner on Phase 2 Intelligence track activation, define ML-specific quality needs
- DevOps / Platform Engineering: Collaborate on CI/CD pipeline templates, infrastructure provisioning for test environments
- Security Team: Integrate security scanning into quality gates, align on SAST/DAST policy
Quarterly Review
No quarterly reviews yet — first review planned for the end of Phase 1 (approximately Month 3).