Designing AI-assisted data stewardship without moving master-data authority
A case about data stewardship, not data cleaning: AI assists; master-data authority does not move.
The case in brief
Current reality
Operational data holds probable duplicate organizations, inconsistent names, missing classifications, incomplete descriptions and conflicting information—some of it linked to ERP identities, open pipeline and reporting—and a proposal to “let AI fix the data”.
What must become true
AI reduces stewardship effort by finding candidates, ranking them by impact, collecting evidence, explaining differences and proposing changes. Accountable people keep every governed master-data decision, and each accepted decision is executed through the established, governed system process.
Design question
Where can AI make data stewardship faster—and which identity and classification decisions remain too consequential to delegate?
01The reality · current state
What happens today—or in the pilot.
Years of manual entry left the CRM with probable duplicate organizations, names in several spellings, accounts without a classification, thin descriptions and relationships that contradict each other. A model appears able to match, normalize, classify and enrich all of it. But every change can touch identity, ERP links, ownership, segmentation, pipeline, reporting and historical continuity—and some of those changes cannot be safely reversed.
- Probable duplicates split pipeline, activity and reporting
- Names are spelled several ways, so searches miss accounts
- Missing classifications leave segments incomplete
- Parent and branch relationships contradict each other
- A model was proposed to merge duplicates above a threshold
- Stewards work from spreadsheets, with no queue and no record of why
What the apparent request hides
“Let AI fix the data” hides the questions that decide what may change—and who is accountable when it does.
- What does “same customer” mean?A duplicate exists only against a defined identity model and key.
- Which fields are authoritative where?A legal name lives in master data; a display name may not.
- What would the change touch?ERP links, ownership, segments, pipeline, reporting, history.
- What can be undone?A merge that moves ERP or legal links is rarely reversible.
- What counts as evidence?Name similarity alone produces noise that erodes trust.
- Who decides—and with what reason?Without a recorded reason, nothing can be learned from a rejection.
02Work decomposition
Reasoning, rules and authority—kept apart.
The work splits by consequence, not by what the model can technically do.
AI reasoning
- Find probable duplicates and name variants
- Compare records and explain what differs
- Suggest a classification with its reason
- Draft a description from cited sources
- Rank suggestions by impact
Deterministic system rules
- Select the records in scope
- Check identity keys and links
- Suppress suggestions below the evidence standard
- Key each task to its records
- Execute accepted changes through the governed process
Human authority
- Whether two records are the same customer
- Merge, and which record survives
- ERP and legal identity links
- Legal names
- Consequential classifications
- The reason for each decision
Reasoning finds and explains. It does not merge, rename, relink or reclassify—no tool the agent can call does any of those.
03Evidence & impact
Enough evidence to decide—or no suggestion at all.
A suggestion reaches the queue only when its evidence meets the standard for its finding type; weak evidence is suppressed, not ranked lower. Evidence is held to the same tests as any context—relevant, trusted, timely, permitted, identified and structured—and the queue is ordered by consequence so scarce steward time goes where a change matters most.
Designing customer and master-data identity across CRM and ERPWhich identities exist, which keys matter, who issues them, how merges and aliases behave and who has authority are decided there—identity architecture comes first. This case does not repeat it.Evidence standard
| Signal | Counts as evidence when | Never enough on its own |
|---|---|---|
| Tax identifier | Identical and valid for both records | When one record is a branch of the other |
| Web domain | Shared and not a public mail provider | For groups that share a domain across legal entities |
| Address | Normalized and identical | For shared offices or industrial estates |
| Name similarity | Combined with another signal | Always—it is a hint, not evidence |
| Order and activity history | Overlapping contacts or products | For distributors serving the same end customers |
Impact
Generic signals, not a scoring formula.
| Signal | Why it raises priority |
|---|---|
| Linked transactional identities | A change can move ERP links and history |
| Open commercial work | Pipeline, quotes and cases would follow the change |
| Active customer | People are working with the record now |
| Dependent records | Contacts, sites and relationships inherit it |
| Reporting and management significance | Segments, targets and reviews read it |
04Agent workflow
Each step with its actor, its tool and its authority.
CRMERPData platformSteward queue
Walk the workflow step by step, or select any step.
01Detect
A scheduled scan selects records by rule: similar names, shared domains, missing fields, conflicting links.
- Who acts
- System
- Authority
- Read
02Suggest
Duplicate candidates, a normalized display name, a probable classification, a drafted description.
- Who acts
- Agent
- Authority
- Recommend
- Tool
Find candidates
03Evidence & impact
Why—shared tax identifier, domain, address, history—what differs, and what the change would touch.
- Who acts
- Agent
- Authority
- Recommend
- Tool
Read record links
04Steward queue
Each suggestion becomes a task keyed to its records, ranked by impact.
- Who acts
- Agent
- Authority
- Execute
- Tool
Create steward task
05Steward decision
Accept, edit or reject—always with a reason.
- Who acts
- Human
- Authority
- Human approval
06Governed action
The system executes the accepted change through the established merge or update process and its controls.
- Who acts
- System
- Authority
- Execute
07Outcome recorded
Decision, reason and result are kept as evaluation evidence.
- Who acts
- System
- Authority
- Read
05Key dimension · Suggest with evidence, steward decides, system applies
Suggestion, decision and action stay three different things
AI works on the left of the boundary; a steward decides on it; a governed system process acts on the right. Nothing in the flow lets the agent edit master data.
- SystemDetect
- Scan by rule
- Records in scope
- AgentAI suggests
- Candidate
- Evidence
- Differences
- Impact
- SystemSteward queue
- Ranked by impact
- One task per candidate
- HumanHuman decision
- Approve
- Edit
- Reject
- Reason
- SystemGoverned system action
- Established process
- Controls and audit
- SystemOutcome feeds evaluation
- Reason codes
- Evaluation cases
- Rule review
- Merge customer identities
- Move ERP or legal identity links
- Change legal names
- Apply consequential classifications
- Overwrite master data
| Finding | AI may suggest | Steward decides | Evidence required | Impact |
|---|---|---|---|---|
| Probable duplicates | A candidate pair with similarity evidence and the differences | Whether they are the same customer, which record survives and what happens to ERP links | Tax identifier, domain, address, contact overlap—and what differs | High when either record has ERP links or open work |
| Display-name normalization | A normalized display name | The display name; the legal name stays as registered | The variants found and their sources | Search and reporting |
| Missing classification | An industry or segment with its reason | Any classification that drives segments, routing or targets | Products bought, public description, similar accounts | Segments and routing |
| Incomplete description | A drafted description | Whether it is accurate enough to save | A cited source for every claim | Low—context for people |
| Inconsistent relationships | The contradiction, with both versions | The correct parent, branch or group link | Both links and where each came from | Ownership, visibility and roll-ups |
AI cannot fix an undefined identity model. It can make the steward’s queue shorter and better evidenced; the key and the decision stay with people.
07Ambiguity policy
When the agent is unsure.
- The identity model does not cover the caseStop
Raised to the identity owner; no suggestion is made
- Candidates linked to different ERP customersEscalate
Marked high impact; steward and finance decide together
- Weak evidence onlyStop
Not suggested—noise erodes steward trust
- Classification sources disagreeContinue
Suggested with both reasons shown
- Record changed since the suggestionRetrieve
Evidence refreshed before the steward sees it
08Human review
Review where ambiguity becomes consequential.
Every suggestion is a review point, because every change it proposes is consequential. The review is designed so a steward can decide quickly and well.
| Review point | Why here | What the person sees | What they can do | What happens next |
|---|---|---|---|---|
| Each suggestion in the queue | Master-data changes affect identity, links, ownership and reporting—and some cannot be undone | The candidate, the evidence, what differs and what the change would touch | Accept, edit or reject with a reason; ask for more evidence | The governed system process applies an accepted change; the decision and reason are recorded |
09Evaluation
Scored per dimension—never one accuracy number.
| Dimension | Question | Measure | Target |
|---|---|---|---|
| Suggestion precisionTask quality | Are suggestions right? | Accepted ÷ suggested, by finding type | ≥ 85 % (illustrative design target) |
| Evidence completenessEvidence & correctness | Does each suggestion show why—and what differs? | Suggestions with complete evidence | 100 % (illustrative design target) |
| Entity correctnessEvidence & correctness | Are the records the ones they appear to be? | Suggestions whose records are correctly identified | ≥ 99 % (illustrative design target) |
| Impact-ranking usefulnessTask quality | Do consequential cases come first? | Accepted high-impact suggestions near the top of the queue | Tracked, no target yet |
| Steward acceptanceTask quality | Is the queue worth working? | Share of the queue decided within its review period | Tracked, no target yet |
| Unapproved master-data changesAuthority compliance | Did anything change without a steward? | Consequential changes without an approved decision | 0 — release gate |
10The trade-offs
Credible options, judged against these premises.
Model merges duplicates above a confidence threshold
None for customer master data
Cost: Irreversible changes across CRM and ERP; confidence becomes permissionAI cleans in bulk, with an undo
Cosmetic fields only
Cost: An undo cannot restore moved ERP links or historyRules-only matching
Clean keys and consistent entry
Cost: Misses the fuzzy cases that matterAI suggests with evidence into a steward queue; stewards decide; the system applies
Messy data with a defined identity model
Cost: Steward capacity, a queue to run and reasons to recordDesign decision
Use AI for evidence-backed candidate generation and prioritization; place suggestions in a governed steward queue; keep identity, merge and consequential classification authority with accountable humans; execute accepted decisions through system-owned processes; and use steward outcomes to improve evaluation and deterministic controls instead of giving the model authority.
AI can help identify inconsistencies. It cannot decide what “same customer” means.
11Audit, governance & learning
Consequential activity can be reconstructed afterwards.
Every accepted change can be traced from the finding to the governed operation—and steward outcomes improve the system without handing the model authority.
- The finding and the records in scope
- The suggestion, its evidence and the record versions it used
- The tools called
- The steward’s decision and reason
- The governed operation applied, with its reference
- The outcome afterwards
Steward outcomes are evidence for people who govern the system, reviewed periodically—not self-training on every click.
- Evaluation datasets
- Matcher analysis
- Taxonomy refinement
- Deterministic rules
- Threshold review
- Controlled prompt and reasoning changes
- Grant the model new permissions
- Train on each decision without review
- Change the identity model
The master-data governance owner, with a periodic review of outcomes and rules
12The second layer
Questions that change the design.
Authority
- Who owns the decision to merge?
- Which fields may never change by suggestion?
- How are ERP-linked records handled?
Evidence & impact
- What evidence does each finding type require?
- How are differences shown?
- Which signals raise priority—and who sets them?
Governance
- How do rejections improve the next scan?
- Who reviews rules and thresholds, and how often?
- How is steward capacity planned?
13Decisions & outputs
What the work produces.
- 01Finding types and evidence standard
- 02Impact signals and steward queue
- 03Authority split
- 04Tool contracts
- 05Reason codes
- 06Evaluation set
- 07Governance review