RIT · QRE
Product-data enrichment
Product visual coming soon
From incomplete catalogue to usable referential: enrichment, classification and attribute-by-attribute validation.
The NEXA impact
82%
first-proposal precision
21,000
references processed per year
To document
processing time saved
The process carries the first proposal across tens of thousands of references — classification, attributes, labels — and brings the team back to a validation role. The quality audit runs continuously: the referential no longer degrades in silence.
This process’s figures will be published with their source and methodology note once the reference set is arbitrated. No unverified number goes live.
Capabilities
Volume processed, validation kept
Incomplete records become filled, classified records. A human still validates record by record, attribute by attribute — the condition for trusting the referential.
A continuous quality audit
Six independent measures score the referential’s health: homogeneity, confusability, classification precision, volumetry, completeness, name cleanliness. Worksites are prioritised by impact and effort.
Enrichment at the source
Supplier files are integrated, matched and normalised upstream. Heterogeneous headers no longer contaminate the referential.
The action plan, not just the finding
The audit leads to actionable worksites: merges, splits, remappings, clean-ups — each with estimated impact and effort. Referential leadership arbitrates on facts.
One family in a structured portfolio.
How the process unfolds
Six steps, six evidence artefacts. That is what separates a critical process from text generation.
Ingestion & context
Sources selected, access scope applied, personal data detected before execution.
Generation
The engine produces a proposal. Every assertion is linked to its source.
Automated check
The AI judge scores compliance, consistency and substantiation, block by block.
Expert validation
Draft → Reviewed → Approved. Three separate roles: operator, reviewer, approver.
Governed feedback
The expert’s feedback is typed, attributed, versioned and replayed on the next process.
Capitalisation
The validated pattern serves the next process, which starts faster and costs less.
Users and data
- Users
- Product marketing, catalogue data, e-commerce teams
- Validators
- Referential owner, product manager
- Data mobilised
- Product referential, supplier files, internal taxonomies
- Integrations
- PIM, warehouses, supplier connectors
Controls and guardrails
- Substantiation threshold configurable per process
- Validation disabled when evidence is insufficient
- Personal-data detection before execution
- Column contract enforced: any gap is rejected
- Runs replayable on an auditor’s request
- Audit Pack exportable over the chosen period
ROI at our clients
MANUTAN
Stake
21,000 references to classify and enrich every year, by hand, with no capacity to keep up with the catalogue.
Goal
Let the platform carry the first proposal and bring the team back to a validation role.
82% first-proposal precision
21,000 references processed per year
Published indicators are orders of magnitude measured in client context, not contractual guarantees. No logo, name or figure is published without the client’s explicit consent.
The trace this process leaves
Transparence
The trace this process leaves
Run Receipt, Evidence Panel in the interface, exportable Audit Pack and replay on demand.
See the evidence chain →Gouvernance
What this process consumes and returns
Cost per run, tokens, CO₂ footprint, number of uses, acceptance rate and ROI configured by the process owner.
See the Cockpit →Put this process into production.
A demo on your own documents, not on a demo dataset.