Case Study - NAPA Auto Parts
NAPA: 96% of reconciliation time eliminated thanks to AI
Case Study - NAPA Auto Parts
The Challenge
The Collaboration
Artificial Intelligence Deployed
The Results
Conclusion
NAPA Auto Parts, Carrefour industriel LDG inc is a NAPA franchise based in Mont-Laurier, Quebec. Like many automotive distributors, the company manages a significant volume of supplier transactions: orders, receipts, invoices. Behind the well-organized counters lay an exhausting administrative process.
Each week, three employees dedicated a total of 60 hours to a single task: manually cross-referencing .SPL files—ephemeral receipt data that disappears after a few days—with supplier PDF invoices. The work was linear and repetitive: import, read, search for purchase order matches, compare quantities and prices, document discrepancies. All done manually.
The problem wasn't limited to wasted time. Errors crept in. Anomalies were often discovered only after payment, leading to difficult-to-resolve supplier disputes. And because everything was undocumented, auditing was impossible. No history, no traceability, no audit trail. As volumes increased, so did the workload. The process didn't scale.
It all began with an AI Strategy (DiagnosticOS), a deep dive into NAPA's operational reality. The objective: to precisely understand the workflow, identify friction points, and determine if automation could absorb the workload without sacrificing human oversight.
The conclusion was clear: 95% of the reconciliation work was mechanical, repeatable, and automatable. The remaining 5%, involving exceptional decisions and judgments on complex anomalies, needed to remain with the team.
In 167 development hours, Quartier AI built NapaOS, custom-tailored for the reality of a franchised automotive distributor. The solution was not generated from a template. It was modeled on NAPA's actual workflow: its suppliers, file formats, purchase order conventions, and tolerance thresholds.
Deployment followed the same logic: first, simple cases to validate system reliability with the team. Then, complex cases, progressively. On the first day, the results were already visible.
The system automatically detects the file type (SCR026 for transaction details, SCR053 for purchase order receipts) and extracts structured data without human intervention.
Every supplier invoice is read automatically, regardless of its format. Purchase order numbers, item references, quantities, and prices are captured and structured in seconds.
Cross-referencing between .SPL files and invoices is done item by item, purchase order by purchase order. 95% of reconciliations are processed without human intervention.
Every price variance, quantity discrepancy, missing item, or surplus is flagged before payment, in a prioritized and actionable list. The team intervenes where necessary, not everywhere.
Every reconciliation, validation, and exception is automatically documented. Supplier audits now have a clear and reproducible trail.
As NAPA adds suppliers or increases its volumes, NapaOS keeps up without a proportional increase in administrative burden.
Since the full deployment of NapaOS, NAPA Auto Parts has eliminated the main source of friction weighing on its operations: reliance on manual reconciliation for every supplier transaction.
The 60 hours per week previously spent cross-referencing .SPL files and PDF invoices, line by line, purchase order by purchase order, have been absorbed by the platform. This time is now dedicated to supplier relations, client development, and growth.
The discrepancy detection rate has reached 100%. Every anomaly is intercepted before payment. Post-payment disputes, which existed precisely because they were difficult to prove, have disappeared from daily operations.
Time-to-value was immediate. From day 1, the first results were visible. The team didn't have to wait for a learning period or a full cycle to see the difference.
The impact goes beyond hours saved. For a franchised automotive distributor, the real value lies elsewhere: the NAPA team no longer just survives the process. They supervise it.
Automotive parts distributors manage growing volumes, multiple suppliers, and non-standardized formats. Manual reconciliation has long been considered an unavoidable constraint. NAPA shows that it isn't.
Does your organization spend hours each week cross-referencing receiving files and supplier invoices, without being able to guarantee that no discrepancies have been missed?