Project Phases
9-month migration timeline from legacy monolith to .NET 8 microservices. Total effective capacity: ~44 man-months (5 engineers × AI 2x multiplier).
Timeline Overview
Shape Up framework: 9 months = 6 cycles × 6 weeks. Each cycle has 1 week shaping, 4 weeks building, 1 week cooldown. AI multiplier is not constant — ×1.0 in Phase 0 (setup) and Phase 3 (hardening), ×2.0 in Phases 1–2 (extraction).
Service Extraction Timeline
Go-Live Order
Travel BookingMonth 3
Event ManagementMonth 4
WorkforceMonth 6
CommunicationsMonth 7
Reporting (CQRS)Month 7
Payment (Frozen → ACL only)Future
Strangler Fig Cutover Process
Day 1–2: 5% traffic → new service
Day 3–4: 25% traffic (if clean)
Day 5: 50% traffic
Day 6–7: 100% traffic (full cutover)
Day 14: Legacy module decommissioned
YARP config rollback < 5 minutes at any point
Capacity Breakdown
| Phase | Duration | Raw MM | Overhead | Net MM | AI × | Effective MM | Services Live |
|---|---|---|---|---|---|---|---|
| Phase 0AI Foundation | M1 | 5.0 | -3.0 | 2.0 | ×1.0 | 2.0 | 0 |
| Phase 1Core Services | M2–4 | 15.0 | -6.0 | 9.0 | ×2.0 | 18.0 | 2 |
| Phase 2Scale Out | M5–7 | 15.0 | -5.5 | 9.5 | ×2.0 | 19.0 | 5 |
| Phase 3Hardening | M8–9 | 10.0 | -3.5 | 6.5 | ×1.0 | 6.5 | 5 |
| Total | 9 months | 45.0 | -18.0 | 27.0 | variable | ~44.0 | 5 (+1 frozen) |
* Phase 0: AI not yet set up → multiplier = 1.0 | Phase 3: Performance/docs work → low AI leverage → 1.0x
Capacity Distribution by Work Package
| Work Package | Man-Months | Assigned | Phase |
|---|---|---|---|
| AI Foundation & Tooling | 3.0 | D1, D5 | 0 |
| Infrastructure / IaC / CI/CD | 4.0 | D5, D1 | 0–1 |
| API Gateway (YARP) | 2.0 | D1 | 0–1 |
| Shared Kernel & Contracts | 1.5 | D1 | 0 |
| Travel Booking Service | 6.0 | D2, D1 | 1 |
| Event Management Service | 5.0 | D3, D2 | 1–2 |
| Workforce Service | 4.0 | D4 | 2 |
| Communications Service | 3.5 | D4, D3 | 2 |
| Reporting Service | 3.5 | D5, D3 | 2–3 |
| Payment ACL | 2.0 | D1, D2 | 1 |
| React Frontend | 5.0 | D5, D4 | 1–3 |
| Testing & QA | 2.5 | All | 1–3 |
| Performance & Hardening | 2.0 | D1, D2 | 3 |
| Documentation & Handoff | 1.0 | All | 3 |
| Total | 45.0 |
Dev Assignments by Month
Team: D1 (Tech Lead), D2 (Senior Backend), D3 (Backend), D4 (Fullstack), D5 (Frontend/DevOps)
| Month | Phase | D1 – Tech Lead | D2 – Sr Backend | D3 – Backend | D4 – Fullstack | D5 – FE/DevOps |
|---|---|---|---|---|---|---|
| M1 | Phase 0 | AI + Gateway scaffold | Legacy Travel analysis | Legacy Event analysis | React design system POC | CI/CD + IaC (Bicep) |
| M2 | Phase 1 | YARP routing + ACL | Travel API (CRUD) | Event API scaffold | Workforce modeling | IaC environments + CDC |
| M3 | Phase 1 | Travel cutover (5%→100%) | Travel tests + bug fixes | Event business logic | React Travel + Event UI | Monitoring dashboard |
| M4 | Phase 1→2 | Event cutover (5%→100%) | Event tests + ACL harden | Event integration tests | Workforce API impl | React Event UI polish |
| M5 | Phase 2 | Architecture review | Comms event triggers | Comms API (email/SMS) | Workforce logic + tests | React Workforce + Reporting |
| M6 | Phase 2 | Workforce cutover | Reporting query APIs | Comms CDC + cutover prep | React Comms + Reporting | Reporting data aggregation |
| M7 | Phase 2→3 | Comms + Reporting cutover | Cross-service integration | Reporting scheduled jobs | React cross-module | Production monitoring |
| M8 | Phase 3 | Performance optimization | Load testing (40K users) | Bug fixes + edge cases | React performance + a11y | Alerting + DR testing |
| M9 | Phase 3 | Architecture docs | API docs (OpenAPI) | Runbooks + playbooks | Design system docs | CI/CD optimization |
Phase Details
Phase 0
AI Foundation
Month 1
2.0 Effective MM
×1.0 AI
Invest the first month in AI toolchain setup, infrastructure scaffolding, and a pilot migration to validate the approach before touching core domains.
✓ AI toolchain deployed (Cursor Pro, Claude Code, CodeRabbit)✓ CI/CD pipeline (GitHub Actions)✓ Infrastructure as Code (Bicep templates)✓ YARP API Gateway configured with Strangler Fig routing✓ Communications service pilot migration (staging)+2 more
4 milestones
Phase 1
Core Services
Month 2–4
18.0 Effective MM
×2.0 AI
Extract the two highest-value core domains — Travel Booking and Event Management — while establishing the Payment ACL bridge to the frozen legacy module.
✓ Travel Booking service extracted and live (Month 3)✓ Event Management service extracted and live (Month 4)✓ Payment ACL bridge operational✓ Per-service Azure SQL databases (Travel, Event)✓ CDC data sync from legacy DB+1 more
Services: Travel Booking, Event Management, Payment (Legacy)
3 milestones
Phase 2
Scale Out
Month 5–7
19.0 Effective MM
×2.0 AI
Extract the remaining three services — Workforce, Communications (production), and Reporting (CQRS) — completing the full microservices decomposition.
✓ Workforce + Allocation service live (Month 6)✓ Communications service promoted to production (Month 7)✓ Reporting CQRS service live (Month 7)✓ React 18 dashboards for Workforce and Reporting✓ Event schema governance enforced+1 more
Services: Workforce + Allocation, Communications, Reporting (CQRS)
3 milestones
Phase 3
Hardening
Month 8–9
6.5 Effective MM
×1.0 AI
Stabilize the platform at scale — load testing for 40K users, security audit, disaster recovery validation, and documenting the Payment migration plan for the next phase.
✓ Load testing simulating 40,000 concurrent users✓ Security audit and penetration testing✓ Disaster recovery validation (failover + restore)✓ Operational documentation and runbooks✓ Payment modernization plan for post-9-months+1 more
3 milestones
Milestones & Go/No-Go Gates
Risk-based milestone tracking — track by risks eliminated, not features delivered.
M0AI Foundation ReadyWeek 4 (Month 1)
"Can we AI-migrate a service?"
Criteria: CI/CD green, Docker local works, AI tools deployed, Communications service on staging
Owner: Tech LeadRisk eliminated: AI approach proven or disproven
M1Travel Booking LiveWeek 8 (Month 3)
"Can we run dual traffic?"
Criteria: 100% traffic via new service, error rate < 0.5%, latency < 200ms p95
Owner: Tech Lead + StakeholderRisk eliminated: Strangler Fig pattern validated
M2Event Management LiveWeek 14 (Month 4)
"Can services talk to each other?"
Criteria: Travel → Payment ACL working, Event Bus messages flowing, calendar integration
Owner: Tech Lead + StakeholderRisk eliminated: Inter-service communication proven
M3Workforce LiveWeek 20 (Month 6)
"Can we handle the data?"
Criteria: CDC from legacy → Reporting working, per-service DBs for Travel + Event + Workforce
Owner: Tech Lead + StakeholderRisk eliminated: Data migration proven
M4Comms + Reporting LiveWeek 28 (Month 7)
"Can we scale?"
Criteria: 5 services running, auto-scaling, load test passing (40K users)
Owner: Tech Lead + StakeholderRisk eliminated: Production readiness
M5Production HardenedWeek 36 (Month 9)
"Can we operate?"
Criteria: Monitoring + alerting + runbooks in place, DR tested, Payment migration plan ready
Owner: Tech Lead + ManagementRisk eliminated: Operational readiness
Go/No-Go Checklist (per service cutover)
☐ All contract tests passing (Pact verified)
☐ Integration tests green
☐ CDC data integrity verified (new DB vs legacy = match)
☐ Error rate < 0.5% on canary (5% traffic, 48h soak)
☐ Latency p95 < 200ms (or within 10% of legacy)
☐ Rollback tested and documented
☐ Monitoring + alerting configured
☐ Stakeholder sign-off
Risk Buffer & Scope Lever
Risk Buffers
Travel more complex than expected
Buffer: 2 weeks Phase 1 buffer | D1 pair-programs with D2
AI tools don't reach 2x
Buffer: Reduce scope: defer Reporting | Measure velocity monthly
Team member out for 1 month
Buffer: Cross-training (secondary domain) | D2↔D3 swap, D4↔D5 swap
Legacy codebase undocumented
Buffer: Phase 0 analysis + AI ingestion | 1M context window for legacy scan
Azure Service Bus learning curve
Buffer: 1 week buffer per phase | POC in Phase 0
Scope Lever (if behind schedule)
MUST HAVE
Travel + Event Mgmt, Travel + Event UI, CI/CD + Docker + ACR, Contract tests, ADRs + Runbooks
SHOULD HAVE
Workforce + Comms, Workforce UI, Full IaC (Bicep), Load testing, Full API docs
NICE TO HAVE
Reporting (CQRS), Dashboard (Reporting), DR automation, Security pen test, Training materials
Rule: If by month 4 there are not 2 services live → reduce Phase 2 scope to 1 service (Workforce only), defer Comms + Reporting.