Home / Services / Validation & review
Stage 3 · Validation & review

The AI never guesses.

Extracted data runs through a validation engine: required fields checked, missing info flagged, contradictions surfaced. Then your operator reviews, edits and approves — with everything logged.

A validation engine, not vibes

After extraction, every RFQ is checked against configurable freight rules. For Ocean FCL, required fields include origin, destination, equipment type and quantity, cargo-ready date and incoterm — with commodity, weight, volume, delivery address, dangerous goods and temperature requirements configurable per customer and workflow.

Each check resolves to a clear verdict: PASS, WARNING, MISSING or CONFLICT.

Missing information

FieldStatus
Origin
Destination
Equipment
Weight
Incoterm
Cargo ready date
Commodity✕ missing
Delivery address✕ missing

2 details missing — commodity, delivery address.

⚠ Conflict detected — human confirmation required

Equipment quantity

Email says
2 × 40HC
Packing list says
3 × 40HC

Quotiveo does not choose one automatically. The RFQ is marked Needs review until a human confirms.

Contradictions surface — safely

Information is compared across every source. When the email and the packing list disagree, the system refuses to resolve it silently. That single design decision is what makes the output safe to quote from.

System confidence blends AI confidence, source quality, validation results and cross-document consistency — and a conflict always resolves to review required.

Clarification drafts, never auto-sends

When details are missing or ambiguous, Quotiveo generates a clarification email covering only the fields in question — commodity, delivery address, whatever's outstanding. Your operator gets Edit and Send. Nothing leaves the building on its own.

Generated draft

Hi Mia, Thank you for your quotation request. Could you please confirm the following details so we can proceed: 1. Commodity description 2. Final delivery address Best regards, Pricing Team

Audit trail entry

{ "field": "weight_kg", "old_value": 12480, "new_value": 12840, "changed_by": "operator_07", "timestamp": "2026-09-10T14:22:11Z" }

Every AI-generated field is editable — and every edit is recorded.

Human editing, fully audited

If the AI says 12,480 kg and your operator corrects it to 12,840 kg, the change is logged with the field, old and new values, who changed it and when. Over 500–1,000 RFQs, those corrections become a proprietary dataset showing exactly where extraction fails — the foundation for getting systematically better.

Pipeline complete

Approve for pricing — with receipts.

One click moves a reviewed RFQ to Ready for pricing, exporting it to your TMS or pricing workflow with its full extraction, validation and audit history attached.

Ingest

Mailbox webhooks, classification and thread tracking. Revisit →

Extract

Structured JSON with confidence and sources. Revisit →

Validate & review

Rules, conflicts, drafts and human approval. You're here.

Put the review loop to the test.

Bring your trickiest RFQs — the ones with conflicting attachments and half-missing details. That's where Quotiveo earns its keep.

Book a pilot