Deliverable Builder Skill

Generates polished, client-ready consulting deliverables from engagement data. Supports 7 deliverable types with standardized formats, quality gates, and professional presentation.

Trigger

When the user invokes /deliverable-builder <client-name> with an optional --type flag specifying the deliverable type. If no type is specified, prompt the user to choose from the available types.

Prerequisites

• Client profile at clients/<client-name>/profile.json
• Relevant engagement data in clients/<client-name>/ subdirectories (discovery, assessment, architecture, tracking)
• Prior skill outputs to source content from

Deliverable Types

Type	Purpose	Typical Audience	Length
final-report	Comprehensive engagement report	Executive sponsor, steering committee	15-30 pages
status-update	Periodic progress report	Project stakeholders	2-5 pages
board-deck	Executive/board presentation	C-suite, board members	10 slides
implementation-guide	Technical how-to for IT teams	Engineering, IT ops	10-20 pages
training-plan	Skills development program	HR, department leads, trainees	5-10 pages
change-management	Organizational change plan	HR, leadership, change agents	8-15 pages
data-strategy	Data governance and quality plan	CDO, data team, compliance	10-15 pages

Steps

1. Load Engagement Context

Read all available engagement data for the client:

clients/<client-name>/
  profile.json                    # Client basics
  discovery/                      # Market research, competitive landscape
  assessment/                     # Current state, AI readiness, stakeholders
  architecture/                   # Solution design, tech selection, costs
  tracking/                       # Status updates, milestones, progress
  deliverables/                   # Previously generated deliverables

Extract key metadata:

• company_name — for headers and references
• engagement_start_date — for timeline context
• primary_contact — for addressing deliverables
• initiative_names — for scoping content

2. Generate Deliverable by Type

Select the appropriate template and populate with engagement data.

---

Type: final-report

Full Engagement Report Structure:

# AI Strategy & Transformation Report
## [Company Name]

**Prepared by:** [Consultant Name / Firm]
**Date:** [YYYY-MM-DD]
**Version:** 1.0
**Classification:** Confidential

---

## Table of Contents
1. Executive Summary
2. Engagement Overview
3. Methodology
4. Current State Assessment
5. Market & Competitive Analysis
6. Strategic Recommendations
7. Solution Architecture
8. Implementation Roadmap
9. Investment & ROI Analysis
10. Risk Management
11. Next Steps
12. Appendices

---

## 1. Executive Summary
[2-3 paragraphs: Context, key findings, top 3 recommendations, expected impact.
This section should stand alone — an executive reading only this page should understand
the engagement's value and recommended path forward.]

## 2. Engagement Overview
- **Objective:** [What was the engagement designed to achieve]
- **Scope:** [What was included/excluded]
- **Timeline:** [Start date] to [End date]
- **Stakeholders Interviewed:** [Count and roles]
- **Methodology:** [Brief description]

## 3. Methodology
[Description of assessment framework, research approach, scoring methodology]

### Assessment Framework
[Description of the maturity model or framework used]

### Research Approach
[How market research, competitive analysis, and trend analysis were conducted]

### Scoring Methodology
[How opportunities were evaluated and prioritized]

## 4. Current State Assessment
### AI Maturity Score: [X/5]
[Pull from assessment/ai-readiness.md]

### Strengths
[Bulleted list of current capabilities and advantages]

### Gaps
[Bulleted list of deficiencies and areas for improvement]

### Technology Landscape
[Summary of current tech stack relevant to AI initiatives]

## 5. Market & Competitive Analysis
[Pull from discovery/market-research.md and discovery/competitive-landscape.md]

### Industry AI Adoption
[Key statistics and benchmarks]

### Competitive Positioning
[Where the client stands vs. peers]

### Key Trends
[Most relevant trends affecting the client]

## 6. Strategic Recommendations
### Initiative 1: [Name]
- **Business Impact:** [HIGH/MEDIUM/LOW]
- **Feasibility:** [HIGH/MEDIUM/LOW]
- **Timeline:** [Months]
- **Investment:** [$Range]
- **Expected ROI:** [Percentage or multiple]
- **Description:** [2-3 sentences]

### Initiative 2: [Name]
[Same structure]

### Initiative 3: [Name]
[Same structure]

### Priority Matrix
| Initiative | Impact | Feasibility | Priority | Phase |
|-----------|--------|-------------|----------|-------|
|           |        |             |          |       |

## 7. Solution Architecture
[Pull from architecture/solution-design.md]
[Include key Mermaid diagrams]

## 8. Implementation Roadmap
[Pull from architecture/implementation-plan.md]
[Include Gantt chart]

### Phase 1: [Name] (Months 1-3)
[Key activities, deliverables, success criteria]

### Phase 2: [Name] (Months 4-6)
[Key activities, deliverables, success criteria]

### Phase 3: [Name] (Months 7-12)
[Key activities, deliverables, success criteria]

## 9. Investment & ROI Analysis
[Pull from architecture/cost-model.md]

### Investment Summary
| Phase | Investment | Expected Return | ROI |
|-------|-----------|-----------------|-----|
|       |           |                 |     |

### 3-Year Financial Projection
[TCO, benefits, net value]

## 10. Risk Management
[Top risks with mitigations]

| Risk | Probability | Impact | Mitigation |
|------|-------------|--------|------------|
|      |             |        |            |

## 11. Next Steps
[Numbered list of specific, time-bound actions with owners]

1. **[Action]** — Owner: [Role], Due: [Date]
2. **[Action]** — Owner: [Role], Due: [Date]

## 12. Appendices
### A. Detailed Assessment Scores
### B. Technology Evaluation Details
### C. Interview Summary
### D. Glossary of Terms
### E. About [Consulting Firm]

---

Type: status-update

Periodic Status Report Structure:

# Status Update: [Company Name]
**Period:** [Start Date] to [End Date]
**Report Date:** [YYYY-MM-DD]
**Prepared by:** [Name]

## Overall Status: [GREEN/YELLOW/RED]

## Period Summary
[2-3 sentence overview of what happened this period]

## Accomplishments
- [x] [Completed item 1]
- [x] [Completed item 2]
- [x] [Completed item 3]

## In Progress
- [ ] [Active item 1] — [% complete, expected completion]
- [ ] [Active item 2] — [% complete, expected completion]

## RAG Status by Workstream

| Workstream | Status | Trend | Notes |
|-----------|--------|-------|-------|
| [Stream 1] | GREEN | Stable | On track |
| [Stream 2] | YELLOW | Declining | [Issue] |
| [Stream 3] | GREEN | Improving | Ahead of schedule |

## Risks & Issues

| # | Type | Description | Impact | Mitigation | Status |
|---|------|-------------|--------|------------|--------|
| 1 | Risk | | | | Open |
| 2 | Issue | | | | In Progress |

## Decisions Needed
1. [Decision needed] — By: [Date], From: [Role]

## Next Period Plan
- [ ] [Planned activity 1]
- [ ] [Planned activity 2]
- [ ] [Planned activity 3]

## Key Metrics

| Metric | Target | Actual | Status |
|--------|--------|--------|--------|
|        |        |        |        |

---

Type: board-deck

10-Slide Board/Executive Presentation Outline:

# Board Presentation: AI Strategy
## [Company Name]
**Date:** [YYYY-MM-DD]

---

### Slide 1: Title
- "AI Strategy & Transformation Roadmap"
- [Company Name]
- [Date]
- [Presenter Name, Title]

### Slide 2: Why AI, Why Now
- Market forces driving AI adoption in [industry]
- Competitive pressure: [2-3 competitor moves]
- Opportunity cost of inaction: [quantified if possible]
- Key stat: "[X]% of [industry] companies have AI in production"

### Slide 3: Where We Are Today
- Current AI maturity score: [X/5]
- Strengths to build on: [2-3 bullets]
- Critical gaps: [2-3 bullets]
- One-line assessment: "[Company] is [behind/on par/ahead] of industry peers"

### Slide 4: Strategic Vision
- Target state in 12-18 months
- Alignment with corporate strategy
- AI principles and guardrails
- North star metric: [the one metric that matters]

### Slide 5: Top 3 Initiatives
| Initiative | Business Impact | Investment | Timeline | ROI |
|-----------|----------------|------------|----------|-----|
| [Init 1] | | | | |
| [Init 2] | | | | |
| [Init 3] | | | | |

### Slide 6: Initiative Deep Dive — [Top Priority]
- Problem being solved
- Proposed solution (high-level architecture diagram)
- Expected outcomes with metrics
- Success criteria

### Slide 7: Implementation Roadmap
- Phase 1 (Q1-Q2): [Headline]
- Phase 2 (Q3-Q4): [Headline]
- Phase 3 (Year 2): [Headline]
- Key milestones and decision gates

### Slide 8: Investment & Returns
- Total investment: $[X] over [Y] years
- Expected annual benefit: $[X] by Year 2
- Break-even: Month [X]
- 3-year ROI: [X]%

### Slide 9: Risks & Mitigations
| Risk | Mitigation |
|------|------------|
| [Top risk 1] | [Mitigation] |
| [Top risk 2] | [Mitigation] |
| [Top risk 3] | [Mitigation] |

### Slide 10: Ask & Next Steps
- **Decision requested:** [What you need the board to approve]
- **Budget approval:** $[X] for Phase 1
- **Timeline:** Begin [Date]
- **Immediate next steps:**
  1. [Action 1]
  2. [Action 2]
  3. [Action 3]

---

Type: implementation-guide

Technical Implementation Guide Structure:

# Implementation Guide: [Initiative Name]
## [Company Name]

**Version:** 1.0
**Date:** [YYYY-MM-DD]
**Audience:** Engineering, IT Operations, DevOps
**Classification:** Internal — Technical

---

## 1. Overview
- Purpose of this guide
- System being implemented
- Architecture reference (link to solution-design.md)

## 2. Prerequisites
### Infrastructure
- [ ] Cloud account provisioned ([provider])
- [ ] VPC/network configured
- [ ] SSL certificates obtained
- [ ] DNS entries configured

### Access & Permissions
- [ ] API keys obtained: [list services]
- [ ] Service accounts created
- [ ] IAM roles configured
- [ ] VPN access for team

### Development Environment
- [ ] Repository access granted
- [ ] Local environment setup documented
- [ ] CI/CD pipeline configured
- [ ] Monitoring tools provisioned

## 3. Architecture Overview
[Mermaid diagram from solution-design.md]

### Component Inventory
| Component | Technology | Version | Purpose | Owner |
|-----------|-----------|---------|---------|-------|
|           |           |         |         |       |

## 4. Step-by-Step Setup

### 4.1 Infrastructure Provisioning
[Detailed steps with commands/configs]

### 4.2 Data Pipeline Setup
[ETL configuration, data source connections]

### 4.3 AI/ML Service Configuration
[Model deployment, API configuration, prompt templates]

### 4.4 Application Deployment
[Deployment procedure, environment variables, health checks]

### 4.5 Integration Configuration
[Connecting to external systems, API mappings]

### 4.6 Monitoring & Alerting
[Dashboard setup, alert rules, escalation procedures]

## 5. Configuration Reference
| Parameter | Description | Default | Required |
|-----------|-------------|---------|----------|
|           |             |         |          |

## 6. Testing Procedures
### Smoke Tests
[Quick validation steps]

### Integration Tests
[Cross-system validation]

### Performance Tests
[Load testing approach and benchmarks]

## 7. Runbook
### Common Operations
[Start/stop, scaling, backup/restore]

### Troubleshooting
| Symptom | Likely Cause | Resolution |
|---------|-------------|------------|
|         |             |            |

### Escalation Path
[Who to contact for different issue types]

## 8. Rollback Procedures
[How to safely revert changes at each stage]

---

Type: training-plan

Skills Development & Training Plan Structure:

# AI Training & Enablement Plan
## [Company Name]

**Date:** [YYYY-MM-DD]
**Version:** 1.0

---

## 1. Skills Assessment

### Current State
| Role Group | AI Literacy | Tool Proficiency | Data Literacy | Gap Level |
|-----------|-------------|-------------------|---------------|-----------|
| Executive | [1-5] | [1-5] | [1-5] | [Low/Med/High] |
| Management | [1-5] | [1-5] | [1-5] | [Low/Med/High] |
| Technical | [1-5] | [1-5] | [1-5] | [Low/Med/High] |
| Operations | [1-5] | [1-5] | [1-5] | [Low/Med/High] |
| Customer-Facing | [1-5] | [1-5] | [1-5] | [Low/Med/High] |

### Target State
[Desired proficiency levels per role group aligned to AI initiatives]

## 2. Training Curriculum

### Track A: AI Awareness (All Staff)
| Module | Duration | Format | Content |
|--------|----------|--------|---------|
| AI Fundamentals | 2 hours | Workshop | What AI is, what it can/can't do, company AI vision |
| AI Ethics & Safety | 1 hour | E-learning | Responsible AI use, bias awareness, data privacy |
| Tool Introduction | 2 hours | Hands-on | Demo of AI tools being deployed |
| AI in Your Role | 1 hour | Workshop | Role-specific AI use cases and workflows |

### Track B: AI Power Users (Department Leads, Analysts)
| Module | Duration | Format | Content |
|--------|----------|--------|---------|
| Prompt Engineering | 4 hours | Workshop | Effective prompting, templates, evaluation |
| Data Preparation | 3 hours | Hands-on | Data quality, formatting, pipeline basics |
| AI Tool Mastery | 4 hours | Lab | Advanced features, integrations, automation |
| Measuring AI Impact | 2 hours | Workshop | KPIs, A/B testing, ROI tracking |

### Track C: AI Technical Team
| Module | Duration | Format | Content |
|--------|----------|--------|---------|
| ML/AI Architecture | 8 hours | Workshop | System design, patterns, best practices |
| Model Operations | 4 hours | Lab | Deployment, monitoring, versioning |
| Data Engineering for AI | 6 hours | Lab | Pipelines, feature stores, vector DBs |
| AI Security & Governance | 4 hours | Workshop | Security, compliance, audit trails |

## 3. Training Timeline

| Month | Track A | Track B | Track C |
|-------|---------|---------|---------|
| 1 | AI Fundamentals, Ethics | Prompt Engineering | ML Architecture |
| 2 | Tool Introduction | Data Preparation | Model Operations |
| 3 | AI in Your Role | AI Tool Mastery | Data Engineering |
| 4 | Refresher | Measuring Impact | AI Security |
| 5-6 | | Advanced workshops | Ongoing labs |

## 4. Resources Required

| Resource | Quantity | Cost | Notes |
|----------|---------|------|-------|
| External trainer (AI) | [X] days | $[X] | Workshops for Track B/C |
| E-learning platform | 1 license | $[X]/yr | Track A modules |
| Lab environment | [X] seats | $[X]/mo | Hands-on practice |
| Training materials | Custom | $[X] | Role-specific guides |

## 5. Success Metrics

| Metric | Target | Measurement Method | Timeline |
|--------|--------|-------------------|----------|
| Training completion rate | >90% | LMS tracking | Month 3 |
| AI tool adoption rate | >70% | Usage analytics | Month 6 |
| Confidence score (survey) | >4/5 | Pre/post survey | Month 3 |
| Productivity improvement | >15% | Process metrics | Month 6 |

## 6. Ongoing Enablement
- Monthly AI lunch-and-learn sessions
- Internal AI champions network
- Quarterly skills reassessment
- Knowledge base of prompts, templates, and best practices

---

Type: change-management

Organizational Change Management Plan Structure:

# Change Management Plan: AI Transformation
## [Company Name]

**Date:** [YYYY-MM-DD]
**Version:** 1.0

---

## 1. Change Overview
- **What is changing:** [Description of the AI initiative]
- **Who is affected:** [Roles, departments, headcount]
- **Why the change:** [Business drivers]
- **Timeline:** [Start to steady state]
- **Scale of change:** [Incremental/Significant/Transformational]

## 2. Stakeholder Analysis

| Stakeholder Group | Impact Level | Current Sentiment | Desired Sentiment | Strategy |
|-------------------|-------------|-------------------|-------------------|----------|
| C-Suite | Medium | Supportive | Champion | Executive briefings |
| IT Leadership | High | Cautious | Supportive | Technical deep dives |
| Middle Managers | High | Concerned | Engaged | Role clarity, training |
| Front-line Staff | High | Anxious | Confident | Hands-on demos, support |
| IT Team | High | Excited | Empowered | Skills development |
| Customers | Low | Unaware | Positive | Gradual rollout |

### Stakeholder Influence-Impact Matrix

      HIGH INFLUENCE
           │

Monitor │ Manage Closely Closely │ (C-Suite, IT Lead) │ ─────────────┼─────────────── HIGH IMPACT │ Keep │ Keep Informed Informed │ (Managers, Staff) │ LOW INFLUENCE

## 3. Communication Plan

| Audience | Message | Channel | Frequency | Owner | Start Date |
|----------|---------|---------|-----------|-------|------------|
| All staff | AI vision & why | Town hall | Once | CEO | Week 1 |
| Managers | Role impact & support | Workshop | Bi-weekly | HR | Week 2 |
| Affected teams | Tool training | Hands-on lab | Weekly | IT | Week 4 |
| All staff | Progress updates | Email/Slack | Monthly | PM | Month 2 |
| Executives | ROI tracking | Dashboard | Monthly | PM | Month 3 |

### Key Messages by Phase

**Awareness Phase (Weeks 1-4):**
- "We are investing in AI to [business objective], not to replace roles"
- "AI will handle [routine tasks] so you can focus on [value-add work]"
- "Training and support will be provided every step of the way"

**Desire Phase (Weeks 4-8):**
- "Early results show [specific positive outcome]"
- "Here's how [role] benefits from the new AI tools"
- "Your feedback is shaping how we roll this out"

**Knowledge Phase (Weeks 8-16):**
- "Here's your training schedule and resources"
- "AI champions in your department can help"
- "Practice environment available for you to explore"

**Ability Phase (Weeks 16-24):**
- "Go-live support is available [channels]"
- "It's normal to feel slower at first — proficiency comes with practice"
- "Share your wins in [channel]"

**Reinforcement Phase (Ongoing):**
- "Celebrating [team/person] for [AI-driven achievement]"
- "Monthly metrics show [positive trend]"
- "New advanced features available for power users"

## 4. Resistance Management

### Anticipated Resistance

| Source | Concern | Evidence | Response Strategy |
|--------|---------|----------|-------------------|
| Middle managers | Job security | Survey data | Role evolution roadmap, upskilling plan |
| Veteran staff | "AI can't do what I do" | Informal feedback | Showcase AI as augmentation, pilot with champions |
| IT team | More work, less credit | 1:1 interviews | Clear ownership, recognition program |
| Compliance | Risk and liability | Policy review | Governance framework, audit trails |

### Resistance Response Framework

1. **Listen:** Acknowledge concerns without dismissing them
2. **Educate:** Provide facts and examples relevant to their role
3. **Involve:** Give resistors a role in shaping the rollout
4. **Support:** Provide extra training and 1:1 coaching
5. **Reinforce:** Celebrate early wins and share success stories

## 5. Adoption Metrics

### Leading Indicators (Predict Success)
| Metric | Target | Measurement | Frequency |
|--------|--------|-------------|-----------|
| Training completion | >90% | LMS data | Weekly |
| Tool login rate | >70% | Usage analytics | Weekly |
| Support ticket volume | Decreasing | Help desk data | Weekly |
| Champion referrals | >2 per team | Champion reports | Bi-weekly |

### Lagging Indicators (Confirm Success)
| Metric | Target | Measurement | Frequency |
|--------|--------|-------------|-----------|
| Process efficiency | +20% | Process metrics | Monthly |
| User satisfaction | >4/5 | Survey | Monthly |
| Error reduction | -30% | Quality metrics | Monthly |
| Full adoption rate | >80% | Usage analytics | Quarterly |

### Adoption Curve Tracking

100% ─────────────────────────────────────── Target 90% ───────────────────────────────── ─ ─ ─ 80% ──────────────────────────── / 70% ─────────────────────── / 60% ────────────────── / 50% ────────────── / 40% ────────── / 30% ───── / 20% ── / 10% / 0% ├──────┬──────┬──────┬──────┬──────┬────── M1 M2 M3 M4 M5 M6 Pilot Early Early Late Late Steady Adopt Major Major Adopt State

## 6. Support Structure

### AI Champions Network
- 1 champion per 15-20 affected staff
- Weekly champion sync meetings
- Champions receive advanced training first
- Recognition and incentive program

### Support Channels
| Channel | Purpose | Availability | Response SLA |
|---------|---------|-------------|-------------|
| Slack #ai-help | Quick questions | 24/7 (async) | 4 hours |
| Weekly office hours | Live Q&A | Tue/Thu 2-3pm | Real-time |
| 1:1 coaching | Individual support | By appointment | 48 hours |
| Knowledge base | Self-service docs | 24/7 | N/A |

---

Type: data-strategy

Data Strategy & Governance Plan Structure:

# Data Strategy for AI
## [Company Name]

**Date:** [YYYY-MM-DD]
**Version:** 1.0

---

## 1. Current State Assessment

### Data Landscape
| Data Source | Type | Volume | Quality (1-5) | Accessibility | Owner |
|-----------|------|--------|---------------|---------------|-------|
|           |      |        |               |               |       |

### Data Maturity Score: [X/5]

| Dimension | Score | Notes |
|-----------|-------|-------|
| Data Quality | [1-5] | [Assessment] |
| Data Governance | [1-5] | [Assessment] |
| Data Architecture | [1-5] | [Assessment] |
| Data Literacy | [1-5] | [Assessment] |
| Data Culture | [1-5] | [Assessment] |

### Current Pain Points
1. [Pain point with specific impact]
2. [Pain point with specific impact]
3. [Pain point with specific impact]

## 2. Target State

### Data Architecture Vision

```mermaid
flowchart TB
    subgraph Sources["Data Sources"]
        Internal[Internal Systems]
        External[External Data]
        Streaming[Real-time Streams]
    end

    subgraph Platform["Data Platform"]
        Lake[(Data Lake)]
        Warehouse[(Data Warehouse)]
        Feature[(Feature Store)]
        Vector[(Vector Store)]
        Catalog[Data Catalog]
    end

    subgraph Consumption["Data Consumption"]
        Analytics[Analytics/BI]
        ML[ML/AI Models]
        Apps[Applications]
        APIs[Data APIs]
    end

    Sources --> Platform --> Consumption

Target State Attributes

• Single source of truth for key business entities
• AI-ready data pipelines with <[X] hour latency
• Automated quality monitoring and alerting
• Self-service data access with governance controls
• Unified data catalog with lineage tracking

3. Data Governance Framework

Governance Structure

Role	Responsibility	Person/Team
Data Owner	Accountability for data domain	[Department heads]
Data Steward	Day-to-day data quality	[Domain experts]
Data Engineer	Pipeline development & maintenance	[IT/Engineering]
Data Protection Officer	Privacy & compliance	[Legal/Compliance]

Data Classification

Classification	Description	Access Level	Examples
Public	Non-sensitive, externally shareable	Open	Marketing materials
Internal	Business data, not sensitive	Authenticated users	Reports, metrics
Confidential	Sensitive business data	Role-based access	Financials, strategy
Restricted	PII, regulated data	Need-to-know	Customer data, health records

Data Lifecycle Policies

Stage	Policy	Retention	Archival
Collection	Consent, minimization	N/A	N/A
Storage	Encryption, access control	By classification	Cold storage at [X] months
Processing	Audit trail, quality checks	During processing	Logs retained [X] months
Sharing	Access approval, anonymization	Per agreement	N/A
Deletion	Right to erasure, verification	Per regulation	Confirmation records

4. Data Quality Standards

Quality Dimensions

Dimension	Definition	Target	Measurement
Completeness	Required fields populated	>98%	Automated checks
Accuracy	Values match reality	>95%	Sampling + validation
Timeliness	Data available when needed	<[X] hours	Pipeline monitoring
Consistency	Same data, same answer	>99%	Cross-source reconciliation
Uniqueness	No duplicate records	>99%	Deduplication checks

Quality Monitoring

• Automated quality scores computed daily per data source
• Quality dashboard visible to data stewards and owners
• Alert thresholds trigger investigation and remediation
• Monthly data quality review meetings
• Quality trends tracked over time

Data Quality Remediation Process

1. Detect: Automated monitoring flags quality issue
2. Assess: Data steward evaluates severity and impact
3. Root Cause: Identify source of quality degradation
4. Fix: Apply correction at source (not downstream patches)
5. Prevent: Add validation rule to prevent recurrence
6. Report: Update quality metrics and stakeholders

5. Data Migration Plan

Migration Scope

Source System	Target	Data Volume	Complexity	Priority

Migration Approach

1. Assess: Profile source data, identify quality issues
2. Map: Define source-to-target field mappings and transformations
3. Cleanse: Apply quality rules, deduplicate, standardize
4. Migrate: Execute migration (batch or incremental)
5. Validate: Reconcile counts, checksums, sample verification
6. Cutover: Switch production to new data platform
7. Decommission: Retire legacy data stores (after validation period)

Migration Timeline

Phase	Duration	Activities	Success Criteria
Assessment	2 weeks	Data profiling, mapping	Complete data inventory
Pilot	2 weeks	Migrate 1 data source	<1% data loss
Wave 1	4 weeks	Core business data	Reconciliation passes
Wave 2	4 weeks	Secondary data sources	All sources migrated
Validation	2 weeks	End-to-end testing	UAT sign-off

6. AI-Specific Data Requirements

Data Requirements per AI Initiative

Initiative	Data Needed	Current State	Gap	Action

Vector/Embedding Data Pipeline

• Document ingestion and chunking strategy
• Embedding model selection and versioning
• Vector store indexing and refresh cadence
• Metadata enrichment for retrieval quality

Training Data Governance

• Labeling standards and quality assurance
• Bias detection and mitigation procedures
• Version control for training datasets
• Reproducibility requirements

7. Implementation Roadmap

Quarter	Focus	Key Deliverables
Q1	Foundation	Data catalog, governance charter, quality baselines
Q2	Pipeline	Core data pipelines, quality monitoring, first migration wave
Q3	AI-Ready	Feature store, vector store, training data pipelines
Q4	Optimize	Self-service access, advanced analytics, quality automation

---

### 3. Apply Quality Gates

Before finalizing any deliverable, validate against these quality gates:

```markdown
## Quality Gate Checklist

### Structure
- [ ] Version header present (Version: X.X)
- [ ] Date stamp present (YYYY-MM-DD format)
- [ ] Executive summary included for documents >50 lines
- [ ] Table of contents for documents >100 lines

### Content
- [ ] All data points have source citations
- [ ] All Mermaid diagrams render without syntax errors
- [ ] Specific action items (not vague recommendations)
- [ ] Each recommendation has: what, who, when, expected outcome
- [ ] Numbers are realistic and sourced (not aspirational)

### Formatting
- [ ] Consistent heading hierarchy
- [ ] Tables properly aligned
- [ ] No placeholder text remaining ([brackets] filled in)
- [ ] Professional tone throughout
- [ ] Spelling and grammar checked

### Completeness
- [ ] All sections populated (no empty sections)
- [ ] Cross-references to other deliverables are valid
- [ ] Appendices referenced in main text

4. Generate Output

Write the completed deliverable to clients/<client-name>/deliverables/<type>.md.

If a file already exists at that path:

1. Check the existing version number
2. Increment the version (e.g., 1.0 -> 1.1)
3. Add a change log entry at the top noting what changed
4. Save as the new version (overwrite with higher version number)

Confirm output by:

• Reporting the file path
• Showing the word count
• Listing the sections generated
• Noting any quality gate items that need manual review