Post-load descartes data validation: shipment count parity, customs filing parity, EDI message control-number coverage, document image hash match rate, freight charge sum reconciliation. Signed evidence pack accepted by internal audit and external SOX/CBP/EU/FDA auditors.
Every Descartes migration produces a load report showing 'success'. But success at the load layer is not the same as data parity at the business layer. Descartes data validation is what proves parity — and it's the only evidence that holds up in an internal audit conversation.
A consultant-led Descartes migration typically validates with spot checks: pull 20 random shipments from Descartes, find them in Fusion, declare success. The problem is the 0.03% of shipments that didn't load — the niche carrier with a malformed lane code, the customs entry whose HTS classification didn't crosswalk, the document image whose hash didn't match after transfer — never surface until a customer dispute, a CBP audit or a freight claim hits in month three of production.
Syntra ETL's descartes data validation reconciles at every level — count, sum and hash — across every shipment, BOL, customs filing, EDI message, document image and rate card. The reconciliation engine runs in parallel across business units and data domains, completing a multi-TB Descartes tenant reconciliation in 24–48 hours. Variances are surfaced with the exact source-system record ID and failure reason — not buried in a 4-million-row CSV that nobody reads.
The output is a signed PDF evidence pack with cryptographic hash signature. Internal audit signs off on the pack directly. External auditors accept it as evidence of migration completeness for SOX 7-year, CBP 5-year post-entry, EU customs 10-year and FDA food-import 2-year retention. The descartes data validation evidence pack is what makes the cutover sign-off conversation a 30-minute review instead of a 3-week negotiation.
Count parity catches missing records. Sum parity catches malformed values. Hash parity catches forensic-level corruption. All three run on every migration.
Shipments, BOLs, customs filings, EDI messages, document images counted at source and target per BU per period. Any variance — even single-digit — surfaces with exact record IDs for investigation. 99.97% isn't acceptable; 100% with documented exclusions is.
Freight charges summed per carrier per period (Descartes vs Fusion), declared value summed per customs entry, rate-card lane sums validated. Catches the malformed cent value, the truncated charge, the unit-conversion error that count parity misses.
Every shipment record hashed (SHA-256) at extraction and re-hashed at load. Every document image hashed at Descartes Document Services extract and re-hashed after binding to Fusion attachment. Hash mismatches caught immediately — partial transfers, encoding corruption, API truncation.
Every CBP entry number, ISF transaction ID and bond reference from Customs Info validated as a cross-reference attribute on the corresponding Fusion GTM Trade Transaction. Missing cross-references flag CBP audit substantiation risk.
ISA/GS control number ranges per trading partner per day validated for 100% capture in the descartes data archive. Gaps (a missed control number range) flagged for re-extraction before sign-off.
Every migrated rate card test-applied to sample shipments; freight calculation result compared to Descartes-computed freight charge. Discrepancies flag translation errors in accessorial-charge rules or fuel-surcharge formulas before cutover.
Continuous validation through the load phase, with a final 48-hour reconciliation pass before cutover sign-off.
Every FBDI load batch reconciled at count and sum level within 1 hour of completion. Variances surfaced to the team daily for investigation. Hash validation runs in parallel on document image batches.
Full historical reconciliation pass covering all loaded shipments, customs filings, EDI messages and document images. Variances investigated and resolved (re-extraction, mapping correction, exclusion documentation).
During the parallel-run window (1–2 shipment weeks of Descartes + Fusion running side-by-side), every delta replay is reconciled to the cent. Freight billing, AP charge reconciliation, customs duty reconciliation.
Complete reconciliation across all domains: shipments, BOLs, customs filings, EDI messages, document images, rate cards. Output: signed PDF evidence pack with cryptographic hash signature ready for cutover sign-off meeting.
Random-sample spot validation for 30 days post-cutover to catch any latent issues. Production support team trained on the descartes data validation evidence pack so post-cutover audit queries resolve in minutes.
A signed PDF with cryptographic hash signature. Internal audit signs off directly; external auditors accept as SOX/CBP/EU/FDA substantiation.
Shipments per BU per fiscal period: Descartes count, Fusion count, variance, documented exclusions. 100% reconciled or explicitly documented.
Freight charges summed per carrier per period: Descartes total, Fusion total, variance in dollars and percent. Reconciled to the cent.
ACE/ISF/ACI/FDA/C-TPAT record counts per filing category with CBP entry number cross-reference confirmation. CBP 5-year retention substantiated.
ISA/GS control number ranges per trading partner per day: Descartes range, archive range, capture percent. 100% capture confirmed per partner.
Per-business-unit hash match rate. 100% match threshold. Hash mismatches re-extracted; final log shows 100% post-re-extraction.
Every record that failed validation with source-system ID, failure reason, resolution action and final disposition. Closes out before sign-off.
Descartes data validation is the post-load reconciliation that proves every shipment, BOL, customs filing, EDI message and document image extracted from Descartes is correctly represented in Oracle Fusion (or in the descartes data archive). Reconciliation runs at three levels: count parity (shipments in Descartes == shipments in Fusion, customs filings in Descartes == GTM Trade Transactions in Fusion, document images in Descartes == attachments in Fusion), sum parity (freight charges per BU per period, declared value per customs entry, carrier rate sum per lane), and hash parity (every shipment hashed at source and re-hashed at target, with hash mismatches surfaced for forensic review). The output is a signed evidence pack that internal audit accepts as proof the migration is complete.
Shipment count reconciliation is the headline metric in descartes data validation: every shipment in the Descartes Global Logistics Network must be accounted for in Fusion Order Management, either as a loaded shipment record or as a documented exclusion (closed shipments older than the operational window routed to the descartes data archive). Syntra ETL's reconciliation engine counts shipments by business unit, fiscal period, carrier and lane, compares Descartes source counts to Fusion target counts, and flags any variance for investigation. A 99.97% match isn't acceptable — the reconciliation engine surfaces the missing 0.03% with the exact source-system shipment IDs and the failure reason.
Customs filing reconciliation is the most legally-significant component of descartes data validation because CBP requires customs records retrievable for 5 years post-entry, and any gap exposes the importer to penalty risk. Syntra ETL counts every ACE entry, ISF filing, ACI declaration, FDA Prior Notice and C-TPAT record in Customs Info, counts the corresponding records in Fusion GTM (or in the descartes data archive for retained filings), and validates that every CBP entry number, ISF transaction ID and bond reference survives the migration as a cross-reference attribute. Sum-level reconciliation validates declared value totals and customs duty totals per import period.
EDI message reconciliation in descartes data validation is structural rather than 1:1 because EDI messages aren't loaded into Fusion as historical records — they're archived for audit retention in the descartes data archive, and the trading-partner registry is re-pointed to Fusion B2B Messaging. Syntra ETL counts every EDI 850/810/856/214/944 message in the Descartes Global Logistics Network for the migration scope window, counts the corresponding archived messages in the descartes data archive, validates ISA/GS control number ranges per trading partner per day, and produces an EDI message reconciliation report showing 100% capture per partner. Any gap (a missed control number range) is investigated before sign-off.
Document image hash validation is the forensic layer of descartes data validation: every BOL, customs form, certificate, packing list and commercial invoice is hashed (SHA-256) at extraction from Descartes Document Services, and re-hashed after binding to Fusion attachments or after landing in the descartes data archive. Hash mismatches surface immediately — caused by partial transfers, encoding corruption or rare API truncation issues — and the document is re-extracted. The hash-validation report shows per-business-unit hash match rate; 100% match is the sign-off threshold. The signed hash log is the legally-defensible evidence of document image integrity for CBP audit substantiation.
The descartes data validation evidence pack is a signed PDF (with cryptographic hash signature) containing: shipment count parity by BU and period (Descartes source vs Fusion target), customs filing count parity per filing category, EDI message control-number coverage per trading partner, document image hash match rate per BU, freight charge sum reconciliation per carrier per period, declared value sum reconciliation per customs entry, carrier rate sum per lane, and the exception register listing every record that failed validation with the specific failure reason. Internal audit signs off on the pack directly. External auditors accept the pack as evidence of migration completeness for SOX, CBP, EU customs and FDA substantiation.
Rate-card parity validation is a specialized subset of descartes data validation: every active carrier rate card in the Descartes Global Logistics Network, Aljex and ShipRush must be represented in Fusion TMS rate-management with full version history. Syntra ETL walks every rate-card record, validates that lane definitions, base rates, accessorial-charge rules, fuel-surcharge formulas and tier-pricing breakpoints survived the translation, and computes a rate-card hash for both source and target. Hash matches confirm structural integrity. The validation report surfaces any rate card where the migrated version produces a different freight calculation than the source — a critical pre-cutover check to prevent shipment-rating discrepancies in production.
Descartes data validation runs continuously during the load phase (weeks 8–12 of a typical migration) and concludes with a final reconciliation pass during the parallel-run window (weeks 12–14). The final reconciliation pass — covering all shipments, customs filings, EDI messages, document images and rate cards — typically completes in 24–48 hours on a multi-TB Descartes tenant because the reconciliation engine runs in parallel across business units and data domains. The signed evidence pack is issued within 48 hours of the final reconciliation pass and reviewed in the cutover sign-off meeting.
Book a 30-minute discovery call. We'll walk you through a sample descartes data validation evidence pack from a recent migration, show you the reconciliation engine output, and explain how internal audit signs off on it directly.