INTERNAL DOC
Rollout Plan
2-phase adoption strategy for the AI Ops Platform — governance, enablement, PII safety, and compliance tooling. L1 AI triage deferred pending agent stability review.
michael@aiops:~$
load --doc rollout-plan --phases 2 --exclude l1-triage --recalc roi
SCOPE NOTE // L1 TRIAGE EXCLUDED
The L1 AI Triage Console is not included in this rollout. The platform is in an active agent-building phase — introducing automated triage on top of agents that are still changing creates compounding instability and unpredictable outputs. This plan covers governance, enablement, PII safety, and compliance tooling only. L1 triage is documented below as a deferred Phase 3, activated once agents stabilize and baseline AI hygiene is established. ROI figures reflect this reduced scope.
TARGET ADOPTION
85–90%
Governance + enablement
ADJUSTED ROI
$2.8M–$3.5M
No L1 triage
L1 UNLOCK POTENTIAL
+$5M–$7M
When agents stabilize
PHASE 01
Foundation
Months 1 – 3
All engineers — staggered by team
ROC first, then SOC, then client-facing
KEY ACTIVITIES
- Agent Tracker goes live — version control for all deployed Copilot agents
- PII Sanitizer enforced before ticket content hits any AI model
- AI Model Guide + Vendor Stack Pages published and accessible
- 2-hour training sessions (live + recorded) covering acceptable use
- SOC 2 evidence collection begins from day one
ROI RAMP
$800k – $1.2M
Governance + safety layer
PHASE 02
Full Rollout
Months 4 – 6
Remaining engineers + all new hires
Mandate enforced for all new tickets
KEY ACTIVITIES
- Shadow AI Gov Console active — flags unauthorized tools and policy drift
- Acceptable-use policy finalized and signed off by all staff
- Copilot Engine Guide deployed — prompt standards and orchestration docs
- BrightGauge + ConnectWise integration complete
- Monthly wins reports delivered to leadership as ROI evidence
ROI RAMP
$2.8M – $3.5M
Full annualized run-rate
ROI: This Scope vs Full Platform
METRIC
THIS PLAN (No L1)
FULL PLATFORM (With L1)
Timeline
6 months
9 months
Minutes saved per ticket
~5 min
~15 min
Conservative ROI (120 eng)
~$1.9M
~$5.6M
Realistic ROI (120 eng)
~$2.8M
~$8.4M
Aggressive ROI (120 eng)
~$3.5M
~$10.1M
SOC 2 Evidence Starts
Day 1
Day 60
Risk to in-flight agents
None — agents untouched
Higher — triage layer on top of active builds
How ROI Ramps (This Plan)
Month 3End of Phase 1
~$800k – $1.2M realized
Month 6End of Phase 2
$2.8M – $3.5M annualized
Phase 3L1 Triage Added
$8M – $10M+ total unlock
120 engineers · 25 tickets/day · 220 days · $120k loaded cost · 5 min saved/ticket (no L1)
L1 Triage — Deferred, Not Abandoned
PHASE 03 // FUTURE STATE // PENDING AGENT STABILITY
L1 AI Triage Console
L1 triage is the single largest ROI driver in the full platform — responsible for roughly 65–70% of total productivity value. It is excluded here not because it lacks merit, but because introducing automated triage on top of actively changing agents creates compounding instability. The right trigger for Phase 3 is agent stability: once the current build cycle settles and the governance layer from Phases 1 and 2 is in place, L1 activates on a foundation that is already audited, policy-compliant, and clean for SOC 2. The platform L1 deploys on top of in month 9 is significantly more defensible than the one it would have deployed on in month 1.
+$5M – $7M
ADDITIONAL ROI AT ACTIVATION
+10 min/ticket
TIME SAVINGS ADDED BY L1
Low Risk
AFTER AGENTS STABILIZE
"Starting without L1 is not leaving money on the table — it is protecting the money already on the table. A stable governance layer now makes the triage unlock significantly safer and more defensible when the time comes."
ROLLOUT PLAN // AI OPS PLATFORM // KRAWCZYK.CITY