What data does Industrial IQ need?

Industrial IQ works from CSV exports. Required fields vary by engine, and the docs list minimum viable uploads plus best uploads.

Are ROI results final financial outcomes?

No. ROI pages label benchmark, estimate, assumption, limitation, and customer-specific diagnostic requirements.

Does Industrial IQ write back to ERP?

No. Industrial IQ creates evidence-backed diagnostics, reports, actions, and governance review without ERP write-back.

Catalog Health Score | AI2COE Industrial IQ Documentation

Task overview

Catalog Health Score

Measures duplicate exposure, description quality, manufacturer and MPN coverage, supplier aliases, UOM consistency, and obsolete-risk signals.

This guide explains the operating task, source data needed, validation steps, output interpretation, trust boundary, and where to go next.

UploadCSV export or sample dataset

Source FitRequired fields and data quality

AI MatchColumn aliases and confidence

NormalizeCanonical field preparation

DiagnoseEngine logic and evidence

ReportScore, actions, assumptions

RepeatScore history and renewal value

Required data by engine

Use this table before upload.

Engine	Required Data File	Required Fields	Optional Fields	Output Generated	Minimum Viable Upload	Engine-Specific Context
PartsCleanse AI	Material or item master CSV	Description	Material Id, Manufacturer, Mpn, Supplier, Uom, Quantity, Unit Cost, Site	Catalog health score, evidence, report, actions, score history	Description	Material Id, Manufacturer, Mpn, Supplier, Uom, Quantity, Unit Cost, Site

Accepted CSV format

Prepare clean exports before upload.

Use UTF-8 CSV with one header row, one operational record per row, stable identifiers, ISO-style dates where available, numeric values without currency symbols, and separate columns for site, supplier, asset, quantity, value, and owner context.

Common errors: merged header rows, hidden subtotal rows, mixed currencies in one value column, duplicate column names, free-text dates, missing material or asset identifiers, and supplier names embedded inside descriptions.

If fields are missing: Industrial IQ still profiles source fit, but unmapped required fields reduce the AI Match Score, Mapping Readiness Score, diagnostic confidence, and report completeness until the user normalizes the data.

Sample row examples

Minimum viable row patterns.

Catalog Intelligence: Description: 6205-2RS bearing sealed SKF
Inventory Risk Intelligence: Material Id: MAT-100245 | Quantity: 42
Procurement Leakage Intelligence: Po Number: PO_NUMBER | Description: 6205-2RS bearing sealed SKF
Working Capital Intelligence: Material Id: MAT-100245 | Stock Value: 1850.00

Related engines

Which diagnostic should the buyer run next?

Related engine

PartsCleanse AI

MRO catalog deduplication, field quality, UOM consistency, and duplicate capital exposure.

Open documentation

Report anatomy

What the buyer can share internally.

Industrial IQ reports include scope, source-fit score, AI match score, mapping readiness, diagnostic confidence, score components, evidence records, assumptions, limitations, review requirements, action items, and recurring score history.

Catalog Health ScoreValue lever

Uploaded dataCustomer-specific diagnostic required

PartsCleanse AIRelated diagnostic engine

EstimateAssumption-labeled until reviewed

Benchmark assumptionUploaded evidenceConfidence tierHuman reviewReportable value

Documentation context

Use the docs to move from source export to reviewed diagnostic evidence.

Each guide keeps Industrial IQ documentation tied to source preparation, validation, evidence review, and Trust controls.

Docs home

Prepare export Use approved CSV/workbook exports or governed sample data. Map fields Confirm required fields, aliases, source fit, and limitations. Review evidence Inspect findings, confidence tiers, assumptions, and owner actions. Trust controls No ERP write-back, source-file purge after report generation, and human review before action.