+ / − on a role card changes headcount; the total must be 25 (counter turns red otherwise).
✅ Validate team — POST api/team/plan. The backend checks each role against minimums, computes migration throughput (engineers × 35 servers/month × months), checks the QA-to-engineer ratio and returns issues and strengths. Cards turn green (good) or red (bottleneck).
⭐ Load recommended fills the brief's 2/3/8/4/4/4 split. Click a role title to see its dependencies in the war-room flow.
0of 25 allocated
—server capacity in 18 months
—months needed
Validation results appear here.
War-room dependencies
BSAs×4
➜
Architects×2
➜
Mig. Engineers×8
➜
QA×4
➜
SREs×4
PMs×3 · orchestrate
Click a role to see its dependencies.
Deep dive
How
Size from throughput: servers ÷ months ÷ servers-per-engineer-month → engineers.
Keep QA at ≥ 1 per 2–3 engineers.
Architects are few but gate everything early; SREs are few but gate everything late.
Why
A Migration CoE centralises patterns, tooling and lessons so every wave is faster than the last.
The war room is the escalation path during cutovers and hypercare.
When
Mobilise at T-9 months after CFO approval.
Scale SREs up and migration engineers down as waves complete.
Pitfalls
All engineers, no BSAs: you migrate fast into the wrong dependencies.
No SREs: Day-2 chaos, bill shock, nobody owns backups.
One PM for 5,000 servers: CAB paperwork becomes the bottleneck.
Key ideas
The counts in the brief total 25
2 Architects + 3 PMs + 8 Migration Engineers + 4 BSAs + 4 QA + 4 SREs = 25. The playground enforces this total so trade-offs are real: adding a tester means removing someone else.