{ "schema": "yarn-pack/2", "id": "framework-dmbok", "version": "0.1.0", "name": "DAMA-DMBOK (framework self-assessment)", "engagement": "AISG — framework self-assessment: workshop prep, workshop, or diagnostic strand", "intro": "A guided conversation against DAMA-DMBOK, not a form. Answer in your own words and name the document or record that shows it if you can. About 22 minutes. Readiness of the eleven DAMA Wheel knowledge areas (DMBOK2 Revised Edition 2024) to supply the data foundations for AI. Excludes the model lifecycle, fairness, explainability and oversight: DMBOK is silent there.", "tone_default": "professional", "prefill_fields": { "department": [ "Executive", "Finance", "Operations", "Customer / Sales", "Technology / IT", "Data & Analytics", "People & Culture", "Risk & Compliance", "Marketing", "Product" ] }, "scales": [ { "id": "maturity5", "name": "Maturity (distilled model)", "levels": [ { "value": 1, "label": "Does not exist", "gloss": "No capability. Absent, or purely ad hoc / accidental." }, { "value": 2, "label": "Partially exists", "gloss": "Emerging and inconsistent. Pockets of activity, not joined up." }, { "value": 3, "label": "Fully exists", "gloss": "Defined, documented and operating across the organisation." }, { "value": 4, "label": "Fully exists & optimised", "gloss": "Measured, refined and improving against targets." }, { "value": 5, "label": "Fully exists & adaptive", "gloss": "Continuously self-adjusting; a source of advantage." } ], "signals": { "1": [ "no ", "not ", "none", "never", "don't", "do not", "nothing", "absent", "unaware", "haven't", "ad hoc", "ad-hoc", "nonexistent", "no idea", "not really" ], "2": [ "some ", "starting", "beginning", "emerging", "pilot", "trial", "informal", "inconsistent", "pockets", "a bit", "occasionally", "early", "experiment", "trying", "patchy" ], "3": [ "documented", "defined", "standard", "standardised", "established", "policy", "framework", "process", "consistent", "across the", "in place", "formal", "governed", "rolled out" ], "4": [ "measured", "metrics", "optimis", "improving", "kpi", "monitored", "reviewed", "refined", "benchmarked", "targets", "tracked", "mature", "regularly review" ], "5": [ "continuous", "adaptive", "self-", "automated end", "best in class", "best-in-class", "competitive advantage", "industry leading", "always", "real-time monitoring", "feedback loop" ] } } ], "categories": [ { "id": "govern", "name": "Governance hub and protection", "order": 1, "target_default": 3 }, { "id": "trust", "name": "Trust-critical three for AI", "order": 2, "target_default": 3 }, { "id": "foundations", "name": "Rim: structure, storage and movement", "order": 3, "target_default": 3 } ], "audiences": [ { "id": "lead", "name": "Governance / risk lead", "desc": "Owns the policy, the register or the risk framework", "deep_dive_sections": [] }, { "id": "owner", "name": "System or use-case owner", "desc": "Runs an AI system or use case day to day", "deep_dive_sections": [] }, { "id": "exec", "name": "Executive / sponsor", "desc": "Accountable for the outcome, not the mechanics", "deep_dive_sections": [] } ], "sections": [ { "id": "core", "title": "DMBOK elements", "blurb": "Everyone answers these. The data office lead or CDO, the data stewards, the owners of each knowledge area (architecture, security, integration, platform and BI, content), and the owner of the AI governance layer on top.", "optional": false, "questions": [ { "id": "dmbok-gov", "type": "scored_text", "category": "govern", "name": "Data Governance (hub)", "text": "When an AI project needs data it has not used before, who decides whether it can have it, and where would I see that decision written down?", "scale": "maturity5", "scored": true, "star": false, "rubric": { "1": "No data strategy, policy or stewardship role exists; decisions about data feeding AI are made ad hoc by whoever holds the system; no issue log is kept.", "3": "Data strategy, policies and standards are approved; stewardship roles and decision rights are named and filled; an issue log is worked; AI use cases share the glossary.", "5": "Decision rights and standards are revised on evidence from the issue log and from AI use-case intake; stewardship covers new AI data sources before they go live." }, "help": "Evidence that would show it: Approved data strategy and policy set with owners; Stewardship role register or RACI for data decisions; Business glossary referenced by AI use cases; Data issue log with dated entries and closures.", "adaptive": { "allow_probe": true, "allow_skip": false, "max_probes": 1 }, "ai_drafted": false }, { "id": "dmbok-sec", "type": "scored_text", "category": "govern", "name": "Data Security", "text": "If I picked a restricted dataset, could you show me that an AI tool cannot read it, and how you last proved that rather than assumed it?", "scale": "maturity5", "scored": true, "star": false, "rubric": { "1": "No data classification; access to stores an AI can reach is granted on request and never reviewed; there is no audit trail of who or what read which data.", "3": "Data is classified; access controls on every store an AI or its pipeline reaches are applied and tested; audit trails record access and are reviewed on a cadence.", "5": "Access rules are re-tested whenever an AI use case, corpus or tool changes; new sources are classified before indexing; audit findings change the rules." }, "help": "Evidence that would show it: Data classification scheme applied to AI-reachable stores; Access control test results, not just configurations; Audit trail review records; Provenance record for training or retrieval data.", "adaptive": { "allow_probe": true, "allow_skip": false, "max_probes": 1 }, "ai_drafted": false }, { "id": "dmbok-dq", "type": "scored_text", "category": "trust", "name": "Data Quality", "text": "For the data your most important AI use case relies on, what do you measure about its quality, how often, and what happened the last time it failed?", "scale": "maturity5", "scored": true, "star": false, "rubric": { "1": "No quality rules or metrics; data fed to AI is not profiled; fitness for use is assumed from the source system’s reputation.", "3": "Quality rules and metrics are defined for each dataset an AI depends on; profiling runs and results are recorded; failures are logged and fixed by a named owner.", "5": "Quality thresholds are tuned from what AI outputs reveal; profiling runs automatically on new or changed sources and blocks use until the rules pass." }, "help": "Evidence that would show it: Quality rules and DQ metrics for AI input datasets; Profiling results with dates and owners; Remediation log for quality failures.", "adaptive": { "allow_probe": true, "allow_skip": false, "max_probes": 1 }, "ai_drafted": false }, { "id": "dmbok-meta", "type": "scored_text", "category": "trust", "name": "Metadata", "text": "Pick an answer one of your AI systems gave last week. Could you trace the data behind it back to where it came from, and who defines what that field means?", "scale": "maturity5", "scored": true, "star": false, "rubric": { "1": "No catalogue or glossary; lineage of data reaching an AI cannot be traced; definitions live in people’s heads or inside individual reports.", "3": "A catalogue and glossary cover the datasets AI uses; lineage from source to AI input is recorded and current; each definition has an owner.", "5": "Lineage and catalogue entries update as pipelines and corpora change; gaps surface automatically and are assigned before an AI use case ships." }, "help": "Evidence that would show it: Metadata repository or catalogue with AI datasets registered; Lineage record from source to AI input; Business glossary with an owner per term.", "adaptive": { "allow_probe": true, "allow_skip": false, "max_probes": 1 }, "ai_drafted": false }, { "id": "dmbok-mdm", "type": "scored_text", "category": "trust", "name": "Reference and Master Data", "text": "If two of your systems disagree about who a customer is, which one does the AI believe, and who decided that?", "scale": "maturity5", "scored": true, "star": false, "rubric": { "1": "Core entities such as customer, product or site exist in several conflicting versions; code sets differ by system; AI uses whichever copy it reaches.", "3": "Golden records exist for the core entities AI uses; reference code sets are published and owned; AI pipelines source from the authoritative version.", "5": "Matching and survivorship rules are refined from conflicts that AI use surfaces; new entities and codes are onboarded to the hub before AI consumes them." }, "help": "Evidence that would show it: MDM hub or golden record register for core entities; Published reference code sets with owners; Evidence AI pipelines source from the master, not a copy.", "adaptive": { "allow_probe": true, "allow_skip": false, "max_probes": 1 }, "ai_drafted": false }, { "id": "dmbok-arch", "type": "scored_text", "category": "foundations", "name": "Data Architecture", "text": "Is there one picture of how data moves across the organisation, and would the data feeding your AI tools be on it?", "scale": "maturity5", "scored": true, "star": false, "rubric": { "1": "No enterprise data model or documented flows; where data comes from and goes to for AI is discovered project by project.", "3": "An enterprise data model and data flow diagrams are current and owned; AI data sources and targets appear on them; a roadmap is approved and followed.", "5": "The blueprint is updated as AI use cases are intaken; new flows to models or retrieval corpora are added to the architecture before build, not after." }, "help": "Evidence that would show it: Enterprise data model with owner and version; Data flow diagrams showing AI sources and targets; Data architecture roadmap with approval.", "adaptive": { "allow_probe": true, "allow_skip": false, "max_probes": 1 }, "ai_drafted": false }, { "id": "dmbok-model", "type": "scored_text", "category": "foundations", "name": "Data Modelling and Design", "text": "Before your last AI project touched a new data domain, was there a model of that domain to read, or did the team work it out from the tables?", "scale": "maturity5", "scored": true, "star": false, "rubric": { "1": "No conceptual or logical models; physical schemas are the only description; definitions are undocumented for the data AI consumes.", "3": "Conceptual, logical and physical models exist for the domains AI draws on; definitions are recorded and reviewed; models are versioned.", "5": "Models are revised as AI use cases surface new requirements; a definition change flows through to the glossary and the pipelines that use it." }, "help": "Evidence that would show it: Conceptual and logical models for AI-relevant domains; Data definitions with review dates; Model version history.", "adaptive": { "allow_probe": true, "allow_skip": false, "max_probes": 1 }, "ai_drafted": false }, { "id": "dmbok-store", "type": "scored_text", "category": "foundations", "name": "Data Storage and Operations", "text": "If the store behind your main AI use case was lost tonight, how long would it take to recover, and when did you last test that?", "scale": "maturity5", "scored": true, "star": false, "rubric": { "1": "No tested backup or recovery for AI-relevant stores; no SLAs; retention and disposal of data and model artefacts are unmanaged.", "3": "Backup and recovery are tested for the stores AI depends on; SLAs are agreed and met; retention and disposal rules are applied and recorded.", "5": "Lifecycle rules apply automatically to new stores, corpora and model artefacts; SLA breaches trigger review; residency and retention are stated per tool." }, "help": "Evidence that would show it: Backup and recovery test records; SLAs for AI-dependent stores with attainment; Retention and disposal records for data and model artefacts.", "adaptive": { "allow_probe": true, "allow_skip": false, "max_probes": 1 }, "ai_drafted": false }, { "id": "dmbok-integ", "type": "scored_text", "category": "foundations", "name": "Data Integration and Interoperability", "text": "When a source system changes a field, how does that reach the AI systems downstream, and how long does it take?", "scale": "maturity5", "scored": true, "star": false, "rubric": { "1": "Pipelines feeding AI are undocumented and owned by individuals; no interoperability standards; an upstream change is found when an output breaks.", "3": "Pipelines to AI are documented with owners and lineage; interoperability standards are published; failures and upstream changes are logged and handled.", "5": "Pipeline changes and upstream corrections propagate to every AI system grounded on them within a known time; monitoring detects drift in the feeds." }, "help": "Evidence that would show it: Pipeline inventory with owners and lineage; Interoperability standards document; Change or withdrawal propagation records.", "adaptive": { "allow_probe": true, "allow_skip": false, "max_probes": 1 }, "ai_drafted": false }, { "id": "dmbok-content", "type": "scored_text", "category": "foundations", "name": "Documents and Content", "text": "What does your AI answer from, who owns that content, and what happens when a document it relies on is withdrawn or corrected?", "scale": "maturity5", "scored": true, "star": false, "rubric": { "1": "Unstructured content has no owner, retention rule or sensitivity tag; AI is pointed at file shares as found.", "3": "Content used by AI has named owners, a refresh cadence and retention rules; sensitivity is tagged before indexing; withdrawals are propagated.", "5": "Corpus registers update as content moves; withdrawals and corrections reach every grounded system automatically; retention applies to the indexes too." }, "help": "Evidence that would show it: Corpus register with content owners and refresh cadence; Records retention schedule covering AI-indexed content; Sensitivity tagging and exclusion rules.", "adaptive": { "allow_probe": true, "allow_skip": false, "max_probes": 1 }, "ai_drafted": false }, { "id": "dmbok-dwbi", "type": "scored_text", "category": "foundations", "name": "Data Warehousing and Business Intelligence", "text": "When an AI tool and a board report quote the same number, do they come from the same place, and who owns that place?", "scale": "maturity5", "scored": true, "star": false, "rubric": { "1": "No governed warehouse or semantic layer; AI and reporting each build their own extracts; figures differ by consumer.", "3": "A warehouse or lakehouse with an owned semantic layer is the source for reporting and AI; definitions match the glossary; access is governed.", "5": "The semantic layer extends as AI use cases need new measures; conflicting figures are detected and reconciled; usage informs what is curated next." }, "help": "Evidence that would show it: Warehouse or lakehouse design with owner; Semantic layer definitions tied to the glossary; Evidence AI use cases source from the governed layer.", "adaptive": { "allow_probe": true, "allow_skip": false, "max_probes": 1 }, "ai_drafted": false } ] } ], "grids": {}, "outputs": [ "Level per element and per category, gated", "Contested-element view (spread of 2 or more)", "Coverage of evidence: confirmed, stated, inferred", "Where to start, foundations first", "Printable client report" ], "report_defaults": [ "rpt-maturity-standard" ], "playbook": { "sequence": [ "govern", "trust", "foundations" ], "sequence_note": "Walk the seven hexagon components for each area and baseline with the DMBOK maturity assessment. Do not claim AI-readiness above the maturity of the weakest relevant spoke.", "actions": { "govern": { "to_3": [ "Name stewardship roles and decision rights; Roles and Organisation & Culture are the two components where accountability lives or dies", "Classify data and test access on every store an AI can reach; record provenance for training and retrieval data" ], "to_5": [ "Pair DMBOK with an operating model (DCAM, Gartner) for the how, and an assurance cadence for the ‘is it running?’ question" ] }, "trust": { "to_3": [ "Define quality rules, catalogue and lineage, and golden records for the datasets AI uses first; these three carry AI trust" ], "to_5": [ "Feed what AI outputs reveal back into quality thresholds, lineage gaps and matching rules" ] }, "foundations": { "to_3": [ "Document the blueprint, models, pipelines, stores and content AI draws on, with owners and tested recovery" ], "to_5": [ "Update architecture, pipelines and corpus registers at AI use-case intake, before build" ] } } } }