Enterprise Data Platform

Operational Runbook & DAMA-DMBOK Teaching Companion

Document Classification: Internal – Training Use Product: Enterprise Data Platform (codename Pipeline Pulse) Live environment: https://pipeline-pulse-79.lovable.app/ Built with: Lovable (frontend) · OpenAI Codex + Claude Cowork (build pipeline) · TanStack Router · Supabase mirror Document owner: Ramachandran Balasubrahmanian — Data Management Leader Aligned to: DAMA-DMBOK, 2nd Edition (DAMA International, 2017)

---

How to read this runbook. This document serves two audiences at once. As an operational runbook it tells you how to run, navigate, and maintain the platform end to end. As a teaching artifact it maps every screen to the DAMA Data Management Body of Knowledge so students can see governance theory expressed in working software. Audience callouts mark whose content a passage is for:

• 👤 Operator — day-to-day users running jobs, checks, and reviews.
• 🎓 Student — learners connecting the UI to DMBOK concepts.
• 🛠️ Admin — platform owners configuring access, tenants, and integrations.

---

Section 1 — Document Control

1.1 Cover & Ownership

Field	Value
Document title	Enterprise Data Platform — Operational Runbook & DAMA-DMBOK Teaching Companion
Document owner	Ramachandran Balasubrahmanian, Data Management Leader
Classification	Internal – Training Use
Version	1.5 (current)
Status	Released
Date of issue	14 June 2026
Review cadence	Quarterly, aligned to DAMA review cycle
Source environment	https://pipeline-pulse-79.lovable.app/
Capture method	Live screen capture of the running application (29 pages), recorded as animated GIFs with static frames extracted for print

1.2 Distribution List

Recipient group	Purpose
Data Management Office (CDO org)	Ownership, governance, sign-off
Data Stewards & Custodians	Day-to-day operation
Data Engineering & DataOps	Pipeline maintenance & reliability
Internal Audit & Compliance	Evidence and control review
Data Management students / trainees	Teaching and certification preparation

1.3 Version History

The platform evolved across an initial build and five formal reviews from March to June 2026. Each row records what changed and which DAMA-DMBOK knowledge area was most affected.

Version	Date	Author / Reviewer	Summary of changes	Sections affected	DMBOK area impacted
1.0	02 Mar 2026	R. Balasubrahmanian (Author)	Initial build: Overview command centre, Job Manager, Projects, Pipeline Maps, Architecture. Core DataOps surface established.	2, 3, 4.1, 4.7–4.11	Data Storage & Operations; Data Architecture
1.1	24 Mar 2026	L. Mendes (DQ Lead)	Added Metadata-Driven Data Quality Control Center with rule engine, reconciliation, and quarantine workflow.	4.17, 5	Data Quality; Metadata Management
1.2	16 Apr 2026	S. Khan (Integration Lead)	Lineage enhancements (AI impact analysis), Data Contract & Schema Registry, breaking-change prevention.	4.7, 4.12	Data Integration & Interoperability
1.3	11 May 2026	T. Brooks (CISO) / G. Patel (Steward)	Governance tokenization/masking, Access (RBAC/ABAC), Audit Trail anomaly detection, Evidence Hub.	4.5, 4.18, 4.24, 4.27	Data Security; Data Governance
1.4	29 May 2026	N. Osei (AI Gov) / Ethics Board	AI Governance model registry, Data Ethics Board, MDM golden records, Catalog & Glossary harvesting.	4.14, 4.16, 4.19, 4.20	Reference & Master Data; AI & Big Data Governance
1.5	14 Jun 2026	R. Balasubrahmanian (Owner)	DAMA Control Tower, Value Realization, Reliability SLO/SLI, Lifecycle & Retention, Document & Content. Full DMBOK coverage map finalised; runbook published.	4.3, 4.4, 4.13, 4.21, 4.22, 6	All 11 knowledge areas (consolidation)

DMBOK Principle Spotlight — Data management is lifecycle management. Notice how the version history itself is a governance artifact: it records who changed what, when, and why, mapped to a knowledge area. DMBOK treats change control and traceability as first-class governance practices, not afterthoughts.

---

Section 2 — Executive Summary & Architecture

2.1 Purpose of the Platform

The Enterprise Data Platform is a DAMA-DMBOK-aligned control plane for an enterprise data estate. It unifies, in a single web application, the operational and governance surfaces that are usually scattered across a dozen tools: pipeline operations, data quality, lineage, metadata and cataloguing, master/reference data, AI model risk, reliability engineering, security and access, retention, and executive evidence. The running system describes itself as a "DAMA DM-BOK2 aligned control plane for governance, quality, lineage, AI risk, reliability, stewardship, and executive evidence."

All data in the environment is synthetic demo / portfolio-simulation data — figures illustrate the operating model, not real financials or customers.

2.2 Intended Users & Business Value

User	What they get	Business value
👤 Operators / Stewards	One place to run quality checks, work queues, trace lineage, resolve incidents	Faster cycle times; fewer escalations
🛠️ Admins / Platform owners	Tenants, access, integrations, billing, settings	Controlled, auditable operations
Executives (CDO/CIO/VP)	Maturity, risk, value, and quarterly priorities	Evidence-based investment decisions
🎓 Students	A working reference implementation of DMBOK	Concepts made concrete

Representative value signals reported in-app (synthetic): +$2.4M total realized value, 95% incident reduction, 70% faster tenant onboarding, 99.7% PII protection rate, 92% audit readiness.

2.3 High-Level Architecture

▶ Live & animated — click to open

Figure 2.1 — Platform architecture, captured live from the Architecture Viewer. Alt text: Three stacked architecture diagrams — a left-to-right Data Flow (Source → Extract → Transform → Load → BI/Analytics), a C4 Container diagram inside a production VPC, and a Semantic Layer blueprint.

The application is a single-page React frontend (built in Lovable) served over HTTPS with JWT auth, talking to an Express API gateway that fronts an execution engine, a fleet of 24 compute nodes, a Postgres metadata store, and S3 object storage. Identity is issued as short-lived tokens via a Firebase-style auth layer with RBAC. The build pipeline uses OpenAI Codex and Claude Cowork; a Supabase mirror provides shared persistence.

Technology stack at a glance

Layer	Technology
Frontend	React SPA generated in Lovable; TanStack Router; client-side route tree (29 routes)
Build pipeline	OpenAI Codex + Claude Cowork
API / services	Express API gateway → Execution engine
Compute	24 compute nodes (auto-scaled)
Metadata store	Postgres
Object storage	Amazon S3 (lakehouse tiers)
Auth & identity	Firebase-style auth, RBAC + ABAC, short-lived JWTs
Persistence mirror	Supabase
External integrations	Slack, PagerDuty, ServiceNow, Datadog, GitHub, Okta, Snowflake, BigQuery

2.4 DMBOK Coverage Map

The eleven DAMA-DMBOK knowledge areas and the platform components that address each. (Maturity levels in parentheses are the in-app self-assessed scores from the DAMA Control Tower, on a 1–5 scale.)

#	DMBOK Knowledge Area	Primary components in the platform	In-app maturity
1	Data Governance	DAMA Control Tower, Executive Governance, Data Governance (tokenization/policy), Evidence Hub, Stewardship, Audit Trail	L4
2	Data Architecture	Architecture Viewer (Data Flow, C4, Semantic Layer), Pipeline Maps	L4
3	Data Modeling & Design	Modeling Studio (Conceptual / Logical / Physical / Standards)	L3
4	Data Storage & Operations	Job Manager, Projects, Reliability, Billing & Usage, Incidents	L4
5	Data Security	Access (RBAC/ABAC), Governance (masking/tokenization), Audit Trail	L5
6	Data Integration & Interoperability	Pipeline Maps (lineage), Data Contract & Schema Registry, Settings (integrations)	L4
7	Document & Content Management	Document & Content Governance (repos, PII scanner, OCR, legal hold)	L2
8	Reference & Master Data	MDM & Reference Hub (golden records, match/merge, survivorship, code sets)	L3
9	Data Warehousing & BI	Overview, Value Realization, Data Product Marketplace	L4
10	Metadata Management	Catalog & Glossary, DQ Metadata Rules, Pipeline Maps inventory	L3
11	Data Quality	DQ Control Center, Reliability, Stewardship queues	L4
+	Extensions: AI & Big Data Governance, Data Ethics, Lifecycle & Retention, Data Management Maturity	AI Governance, Data Ethics Board, Lifecycle & Retention, DAMA Control Tower	L2–L3

DMBOK Principle Spotlight — The DAMA Wheel. DMBOK arranges the knowledge areas around Data Governance at the hub, with the other ten areas as spokes. This platform is laid out the same way: the Executive Layer and DAMA Control Tower sit at the centre, and the operational spokes (quality, lineage, security, MDM, metadata…) radiate outward. See Section 6 for the full DAMA Wheel discussion.

---

Section 3 — Getting Started / Environment Setup

3.1 Prerequisites

Requirement	Detail
Browser	Any modern desktop browser (Chrome, Edge, Safari, Firefox)
URL	https://pipeline-pulse-79.lovable.app/
Account	Demo personas are pre-provisioned; no sign-up required for the training environment
Data	All data is synthetic — safe for classroom use
Screen	1280 px width or wider recommended (the layout is a fixed left sidebar + content canvas)

3.2 Access, Roles & Permissions

The platform ships with five demo personas spanning the role spectrum. Persona and tenant can be switched from the user menu at the bottom-left of the sidebar (avatar → Switch persona / Switch tenant).

Persona	Role	Clearance	Typical use
Ram Balasubrahmanian	Platform Admin	Restricted	Full admin, all tenants
Gita Patel	Data Steward	Confidential	Stewardship, glossary, approvals
Liam Mendes	Data Engineer	Confidential	Pipelines, jobs, reliability
Aisha Rivera	Business Analyst	Internal	Reporting, products, value
Thomas Brooks	Auditor	Restricted	Evidence, audit trail, read-only review

🛠️ Admin note. Roles drive what each persona can see and do; the Access (RBAC/ABAC) screen (Section 4.24) exposes the role-permission matrix and a live policy-decision simulator. Switching persona is the fastest way to demonstrate least-privilege access in a class.

3.3 First-Time Navigation

The left sidebar groups all 29 screens into five collapsible sections. Learning this map is the single most useful orientation step.

▶ Live & animated — click to open

Figure 3.1 — First-time navigation: the sidebar's five groups and the Overview canvas. Alt text: Application shell with a dark left sidebar listing grouped navigation links and a light content area showing KPI cards and charts.

The five navigation groups

Group	Screens
Executive Layer	Overview · Executive Governance · Value Realization · DAMA Control Tower · DAMA Evidence Hub · Demo Video
Data Platform	Pipeline Maps · Architecture · Projects · Data Product Marketplace · Job Manager · Data Contract Registry
Data Governance & Quality	Data Reliability · Catalog & Glossary · Modeling Studio · MDM & Reference · DQ Control Center · Data Governance
AI, Ethics & Compliance	AI Governance · Data Ethics Board · Document & Content · Lifecycle & Retention · Stewardship
Operations & Admin	Access (RBAC/ABAC) · Billing & Usage · Incidents · Audit Trail · Tenants · Settings

Common top-bar controls (present on every screen): global search ("Search jobs, projects, or incidents…"), the active Tenant chip, a Notifications bell, and Settings. A floating Demo Guide button (bottom-right) launches an in-app walkthrough.

Recommended first session (5 steps)

1. Open the Overview to read platform health at a glance.
2. Open the DAMA Control Tower to see knowledge-area maturity.
3. Open the DQ Control Center and click Run Data Quality Validation to watch the rule engine work end to end.
4. Open Pipeline Maps and click a node for AI impact analysis.
5. Switch persona to Auditor and open the Evidence Hub to see read-only, audit-ready evidence.

---

Section 4 — Component-by-Component Walkthrough

Every screen below follows the same seven-part structure: screenshot/GIF + annotations, Purpose, Why it is needed, Step-by-step usage, DMBOK mapping, Student Insight, and Troubleshooting. Figures are animated GIFs in the HTML and Markdown editions; the Word/PDF editions show a representative frame with a link to the live page.

4.1 Overview — Command Centre (/)

Nav group: Executive Layer · Audience: 👤 Operator · Executives

▶ Live & animated — click to open

Figure 4.1 — Overview command centre (animated tour, top → freshness scoreboard). Alt text: Dashboard with four KPI cards, a 30-day completed-vs-failed line chart, a system-pulse panel, and a per-source freshness table colour-coded healthy/warning/breached.

① What you're seeing

Marker	UI element	What it does
1	KPI row	Total Jobs Run 14,823 (+12%), Success Rate 98.2%, Active Pipelines 47 (3 critical), Avg Execution 4.2 min
2	Job Execution Trend	30-day completed-vs-failed line chart
3	System Pulse	Compute Nodes Online 24/24, Queue Depth 3 waiting, Network I/O 2.4 GB/s
4	Data Freshness Scoreboard	SLA status per source: Salesforce 99.1% ✓, SAP ERP 97.3% ✓, AWS S3 88.2% ⚠, Oracle DB 61% ● Breached
5	Visitor tracking + Start Demo	Live session metrics and the guided-demo launcher

Purpose. A real-time command centre that summarises platform health across all tenants and pipelines on one screen.

Why it's needed. Executives and operators need a single, trustworthy "is everything OK?" view. Surfacing a breached freshness SLA (Oracle, 61%) on the home page turns a latent data-trust problem into a visible, actionable one.

Step-by-step usage. (1) Scan the KPI row for throughput and success rate. (2) Check System Pulse for capacity pressure (queue depth, nodes). (3) Read the Freshness Scoreboard top-to-bottom; anything amber/red is a candidate for the Incidents or Reliability screens. (4) Click Start Demo for the guided tour.

DMBOK mapping. Data Warehousing & BI (executive dashboards), Data Quality (freshness/SLA scoreboard), Data Governance (cross-tenant oversight). Principle (paraphrased): trustworthy BI depends on timely, monitored data — freshness and SLA adherence are quality dimensions, not just ops metrics.

🎓 Student Insight. This screen demonstrates that a dashboard is itself a data product with quality obligations. Reflection: which two freshness states here would you escalate first, and to which other screen in this platform would you route each?

Troubleshooting. Numbers look static → it is synthetic demo data refreshed per session. A breached source → open Data Reliability (4.13) for the SLO/error-budget detail and Incidents (4.26) for any linked RCA.

4.2 Executive Governance Dashboard (/executive)

Nav group: Executive Layer · Audience: Executives (CDO/CIO/VP)

▶ Live & animated — click to open

Figure 4.2 — Executive Governance dashboard (animated tour). Alt text: Risk list with High/Medium severity chips, a numbered next-quarter action list, and an ROI summary totalling $2.4M.

① What you're seeing

Marker	UI element	What it does
1	Top 5 Enterprise Risks	Schema drift (High), Unreviewed high-risk AI models (High), Duplicate customer records (Medium), Catalog adoption under target (Medium), Unstructured PII sprawl (High) — each with a mitigation
2	Top 5 Actions — Next Quarter	Approve DAMA roadmap, stand up AI Ethics Board, promote MDM to prod, fund catalog adoption, finish zero-trust onboarding
3	Governance ROI Summary	Avoided incident cost $1.1M; faster onboarding $680K; compute optimization $184K; audit prep saved $420K; Total $2.4M

Purpose. A CIO/CDO-level portfolio view of data trust, risk, value, and the quarter's priorities.

Why it's needed. Governance competes for funding like any program. Pairing risks with mitigations and value with dollars lets leaders make defensible investment decisions.

Step-by-step usage. (1) Review each risk and its mitigation owner. (2) Confirm the next-quarter actions align to the risks. (3) Use the ROI summary in budget conversations. (4) Drill into any risk via its related operational screen (e.g., schema drift → Contract Registry).

DMBOK mapping. Data Governance (strategy, risk, value). Principle (paraphrased): governance must demonstrate business value and manage risk explicitly; a governance program is steered by leadership with measurable outcomes.

🎓 Student Insight. Risks here map one-to-one onto operational screens elsewhere in the app — governance is not a slide, it is wired to the controls that mitigate it. Reflection: pick one risk and name the exact screen that provides the evidence it is being mitigated.

Troubleshooting. ROI figures are synthetic ("Figures illustrate the operating model — not real financials"). Risks not updating → they are curated portfolio items, not live telemetry.

4.3 Executive Value Realization (/value)

Nav group: Executive Layer · Audience: Executives · 🎓 Student

▶ Live & animated — click to open

Figure 4.3 — Value Realization (animated tour). Alt text: Eight outcome KPI tiles, a before-vs-after table, a risk-reduction-by-DAMA-area bar list, and a business-value-by-domain table.

① What you're seeing

Marker	UI element	What it does
1	Outcome KPIs	Incident Reduction 95%, Onboarding Time −70%, Audit Readiness 92%, PII Protection 99.7%, DQ Improvement +34%, Cost Optimization 75%, SLA 99.9%, Reusable Data Products 7
2	Before vs After	e.g., MTTD 47m→3m, MTTR 6h12m→11m, Pipeline failures/wk 38→2, Cost/1M records $1.84→$0.46
3	Risk Reduction by DAMA Area	Security 62→12, Quality 58→18, AI Gov 71→24, Lifecycle 49→17, Metadata 55→22
4	Business Value by Domain	Customer +28% campaign ROI, Finance −3-day close, Risk +12% detection, Claims $4.2M leakage prevented, Privacy 0 consent incidents

Purpose. Translate platform activity into quantified business outcomes by domain and DAMA area.

Why it's needed. It closes the loop between governance effort and measurable value — the evidence base for continued investment.

Step-by-step usage. (1) Read outcome KPIs for headline movement. (2) Use Before/After to show trajectory. (3) Map risk reduction back to the DAMA areas in the Control Tower. (4) Cite domain value in business reviews.

DMBOK mapping. Data Governance (value realization), Metadata Management and Data Quality (the measured dimensions). Principle (paraphrased): managed data is an enterprise asset; its value should be measured and reported like any asset's return.

🎓 Student Insight. "Before vs After" is the data-management business case in miniature. Reflection: which single before/after metric would most convince a sceptical CFO, and why?

Troubleshooting. All metrics synthetic. Values should reconcile with Reliability (MTTR/MTTD) and DQ (improvement %) — note the cross-screen consistency.

4.4 DAMA-DMBOK Control Tower (/dama)

Nav group: Executive Layer · Audience: 🎓 Student · Governance leads

▶ Live & animated — click to open

Figure 4.4 — DAMA-DMBOK Control Tower (animated tour). Alt text: A grid of knowledge-area tiles colour-graded L1–L5, above a table listing maturity, status, evidence, key risk, next action, owner, and steward per area.

① What you're seeing

Marker	UI element	What it does
1	Summary tiles	Overall Maturity 3.4/5, Implemented 7, Partial 7, Open Gaps 0, Critical Risks 2, Audit Evidence 47 artefacts
2	Knowledge Area Heatmap	L-levels per area: Security L5, Governance/Architecture/Storage/Integration/Warehousing/Quality L4, Modeling/Reference/Metadata/AI L3, Doc & Content / Ethics L2
3	Coverage table	Per area: maturity bar, status, linked module (e.g., /governance, /dq), evidence, key risk, next action, owner, steward, last reviewed
4	Export controls	Export DAMA Maturity Report and JSON

Purpose. A single board that scores the estate against all DAMA-DMBOK knowledge areas and links each to its owner, evidence, and remediation.

Why it's needed. Maturity assessment (DAMA-DMM/DCAM style) is how organisations know where to invest. The heatmap makes gaps unmissable.

Step-by-step usage. (1) Read the heatmap for weak areas (here: Doc & Content L2, Ethics L2). (2) Open the table row to see the key risk and next action. (3) Note the owner/steward accountable. (4) Export the report for the governance council.

DMBOK mapping. All eleven knowledge areas, plus Data Management Maturity Assessment. Principle (paraphrased): DMBOK is organised as knowledge areas governed from a central hub; maturity is assessed per area to prioritise improvement.

DMBOK Principle Spotlight — Maturity is per-area, not global. A single "3.4/5" hides the story; the heatmap shows Security at L5 but Ethics at L2. Improvement programs target the lowest, highest-risk areas first.

🎓 Student Insight. This is the DAMA Wheel rendered as an operations board. Reflection: given two areas at L2 (Document & Content, Ethics), which would you fund first and what evidence would prove progress?

Troubleshooting. Levels are self-assessed demo scores. The linked module column is the fastest way to jump from "what's weak" to "where to fix it."

4.5 DAMA Evidence Hub (/evidence)

Nav group: Executive Layer · Audience: Thomas Brooks (Auditor) · Compliance

▶ Live & animated — click to open

Figure 4.5 — DAMA Evidence Hub (animated tour). Alt text: A filterable evidence table (EV-1001…EV-1014) with DAMA area, control, policy, linked asset, owner, steward, type, risk, status, and per-row export.

① What you're seeing

Marker	UI element	What it does
1	Summary tiles	Evidence Artefacts 14, Controls Passing 7, Controls at Risk 5, Pending Review 2, Audit Readiness 50%, Last Review Q2 2026
2	Evidence table	EV-1001…EV-1014 mapping each control to a DAMA area, policy ID, linked asset, owner, steward, risk, and status (Passed / At Risk / Pending / Remediation)
3	Actions	Create Review Task, Generate Audit Pack, JSON, Export CSV

Purpose. A defensible, exportable evidence store proving each control operates — organised by DAMA knowledge area.

Why it's needed. Audits fail on missing or slow evidence. Pre-assembling evidence (and an "Audit Pack") turns days of scramble into minutes.

Step-by-step usage. (1) Filter by area/risk/status. (2) Open an At Risk item to see the failing control. (3) Create Review Task to assign remediation. (4) Generate Audit Pack / Export CSV for auditors.

DMBOK mapping. Data Governance (control evidence, auditability), touching every area it documents. Principle (paraphrased): governance must be demonstrable; controls require evidence that they are designed and operating effectively.

🎓 Student Insight. Evidence is the bridge between policy (intent) and audit (proof). Reflection: for control EV-1003 (AI model card, At Risk), what single action would move it to Passed, and which screen performs it?

Troubleshooting. Audit readiness at 50% reflects 5 controls at risk + 2 pending — work the queue from the Stewardship screen. Export produces synthetic CSV/JSON.

4.6 Demo Video (/demo)

Nav group: Executive Layer · Audience: 🎓 Student · New users

▶ Live & animated — click to open

Figure 4.6 — Demo Video page. Alt text: A media page with an embedded 1080p MP4 player titled "Pipeline Pulse demo", with Open-raw and Download controls.

① What you're seeing

Marker	UI element	What it does
1	Video card	"Pipeline Pulse demo — 1080p MP4 · AI jobs, lineage, anomaly detection, and monitoring"
2	Controls	Open raw and Download the source file

Purpose. A self-contained, contest-ready walkthrough for onboarding and presentations.

Why it's needed. Video lowers the barrier for first-time users and gives trainers a ready asset.

Step-by-step usage. (1) Play inline. (2) Open raw for full screen. (3) Download to embed in a course module.

DMBOK mapping. Document & Content Management (managed media asset) and training enablement. Principle (paraphrased): content and knowledge assets are governed deliverables with ownership and access, like structured data.

🎓 Student Insight. Even a demo video is a content asset with a lifecycle. Reflection: what retention and access rules should govern an internal product-demo recording?

Troubleshooting. Video won't autoplay → use the play control; some browsers block autoplay. Use Open raw if the embed is blocked by a corporate proxy.

4.7 Pipeline Maps — Lineage (/lineage)

Nav group: Data Platform · Audience: 👤 Operator · 🎓 Student