Owner: Head of Delivery (HoD)
Contributors: Account Lead (client communication) · PM (incident coordination) ·
AI Engineer / AIe / AIc Specialist (incident identification + remediation) ·
CEO (P1 sign-off)
Department: D4 — Delivery
Entity: PrivateAI
Status: v1.0 · Complete
Version: 1.0 | Last updated: 2026-07-18
What this workflow does: Governs how delivery incidents — mistakes, failures,
quality issues, scope disputes, and client complaints — are identified, classified
by severity, escalated to the right level, remediated, and formally closed. Applies
across all delivery tracks (AIe, AIc, AI Project) and all engagement phases.
What triggers it: Any of the following:
What it produces: A logged and closed incident in DEL-06-DOC-01 (Delivery
Mistake & Escalation Log) — with a documented remediation plan, CEO sign-off
for P1 incidents, and a lessons-learned entry fed into DEL-07 and DEL-10.
Why it matters: How a team handles mistakes defines its reputation more than
how it handles success. A structured, fast, and transparent escalation process
protects the client relationship, limits commercial exposure, and turns incidents
into institutional learning.
⚠️ CEO involved — P1 incidents. CEO reviews and signs off on every P1
(client-impacting) incident before the incident is formally closed. No P1
closes without CEO sign-off.
| Level | Name | Definition | Response time | CEO? |
|---|---|---|---|---|
| P1 | Critical — client impact | Active harm to client operations, data breach, major deliverable failure, or reputational damage | Same day | Yes — sign-off before closure |
| P2 | Major — delivery risk | Significant milestone missed, repeated quality failures, client escalation, or specialist performance issue affecting delivery quality | Within 24 hours | No — HoD leads |
| P3 | Minor — internal quality | Single missed deadline, minor quality gap, internal process failure — no direct client impact | Within 48 hours | No — PM leads |
| P4 | Observation | Near-miss, potential risk identified, process improvement opportunity — no current impact | Weekly review | No — PM logs |
Auto-escalation rule: 3 P3 incidents within one calendar month from the
same engagement or team member auto-escalates to P2.
| Role | Department | Responsibility | CEO? |
|---|---|---|---|
| HoD | D4 | Incident triage (all levels) · leads P1 and P2 remediation · signs off P3 closure · monitors P4 log | No |
| Account Lead | D4 | All client-facing communication during incidents · manages client relationship throughout | No |
| PM | D4 | Coordinates incident logistics · Jira updates · P3 and P4 lead · tracks remediation actions | No |
| AI Engineer / AIe / AIc | D4 | Identifies and flags incidents · executes technical remediation | No |
| CEO | D6 | Reviews and signs off every P1 incident before closure | Yes |
Step 1 — Identify and flag the incident
Step 2 — Initial severity triage
P1 Response — Critical (client impact)
Step 3a — Immediate containment
Step 4a — Client notification (P1)
Step 5a — Remediation plan
Step 6a — Execute remediation
Step 7a — CEO sign-off and closure (P1)
P2 Response — Major (delivery risk)
Step 3b — HoD-led assessment
Step 4b — Client communication (P2, if applicable)
Step 5b — Remediation and monitoring
P3 Response — Minor (internal quality)
Step 3c — PM-led resolution
P4 Response — Observation
Step 3d — Log and monitor
Step 8 — Root cause analysis
Step 9 — Feed into DEL-10 and DEL-07
| Tool | How it's used | Agent-ready |
|---|---|---|
| DEL-06-DOC-01 | Delivery Mistake & Escalation Log — primary incident record | Yes — API + MCP |
| Jira | Incident flags on delivery tickets · remediation action tracking | Yes — confirmed |
| Google Workspace | Remediation plans, root cause analysis, client communication records | Yes — API + MCP |
| Google Meet | Incident assessment calls, client communication calls, CEO sign-off review | Yes |
| AWS | Containment actions for infrastructure-related P1 incidents | Yes — confirmed |
| Wiki.js | DEL-07-DOC-01 knowledge entry for lessons learned | Yes — native |
Severity triage (Step 2)
| Condition | Severity | Owner |
|---|---|---|
| Active client harm / data breach / major failure | P1 | HoD → CEO sign-off |
| Milestone missed / repeated quality / client escalation | P2 | HoD leads |
| Single minor issue / internal only | P3 | PM leads |
| Near-miss / observation | P4 | PM logs |
| 3 x P3 from same engagement in one month | Auto-escalate to P2 | HoD |
P1 client notification timing (Step 4a)
| Condition | Action | Owner |
|---|---|---|
| Client already aware | Notify immediately — same business day | Account Lead + HoD |
| Client not yet aware | Notify same business day regardless | Account Lead + HoD |
P2 client notification (Step 4b)
| Condition | Action | Owner |
|---|---|---|
| Client visible impact | Notify — HoD-approved message | Account Lead |
| Internal only — not yet visible | HoD decides — notify or remediate first | HoD |
| Scenario | How to handle | Escalation |
|---|---|---|
| Security or data breach during delivery | Treat as P1 immediately. Contain. Notify Account Lead and HoD simultaneously. Legal Lead and CEO informed within 1 hour. | HoD → CEO → Legal Lead |
| Specialist conduct violation | Remove from client-facing work immediately. HoD leads investigation. Account Lead manages client. DEL-02 replacement trigger if Critical. | HoD → CEO |
| Client raises incident that team was unaware of | Treat as minimum P2. Immediate HoD assessment. Account Lead responds to client within 2 hours. | Account Lead → HoD |
| Incident involves third-party tool or vendor | HoD assesses vendor responsibility. Client communication still REMAP.ai's responsibility. | HoD → Account Lead |
| CEO unavailable for P1 sign-off | HoD documents the incident in full. CEO signs off at first availability. Incident stays open until CEO sign-off received. | HoD — no substitute for CEO sign-off |
| Client threatens legal action | Immediately involve Legal Lead. No further client communication without Legal Lead approval. CEO informed. | Account Lead → HoD → Legal Lead → CEO |
| Metric | Type | Target |
|---|---|---|
| P1 incidents per quarter (count) | Lagging | 0 |
| P2 incidents per quarter (count) | Lagging | ≤ 2 |
| P1 client notified within same business day (%) | Process | 100% |
| P1 CEO sign-off obtained before closure (%) | Process | 100% |
| P3 auto-escalation rule applied correctly (%) | Process | 100% |
| Root cause documented for all P1 and P2 incidents (%) | Process | 100% |
| DEL-10 retrospective includes all P1/P2 from the period (%) | Process | 100% |
| Step | Current state | Suggested automation | Tool | Effort |
|---|---|---|---|---|
| Step 1 — Incident flag | Manual | n8n: Jira flag of severity type → auto-create DEL-06-DOC-01 entry and notify HoD | n8n + Jira | Medium |
| Step 2 — Auto-escalation rule | Manual count | n8n: counts P3 flags per engagement per month → auto-upgrades to P2 if threshold reached | n8n + Jira | Low |
| Step 7a — CEO sign-off reminder | Manual | n8n: P1 incident open >24h without CEO sign-off → reminder to HoD | n8n + Jira | Low |
| Step 9 — DEL-10 feed | Manual | n8n: incident closed → auto-creates DEL-10 retrospective agenda item | n8n + Jira | Low |
Upstream (feeds incidents into DEL-06):
Downstream (DEL-06 feeds into):
| Version | Date | Description | Author |
|---|---|---|---|
| 1.0 | 2026-07-18 | Initial release — P1–P4 severity framework, auto-escalation rule (3x P3 → P2), CEO mandatory sign-off on P1 closure, phase-based response paths per severity level, root cause analysis requirement, DEL-10 and DEL-07 feed. | Hamza Tasneem (HoD) |