The Problem
The company invented the modern one-stop shopping concept in 1962. Today it operates 259 supercenters across Michigan, Illinois, Indiana, Kentucky, Ohio, and Wisconsin — the 14th-largest private company in the United States and 23rd-largest retailer by revenue, generating $19.6B in annual sales with 70,000 employees. Its $54M inbound freight spend moved through Intelligent Audit (IA) — a platform the discovery process found had no AI capability, offline rate management in Excel and Transporeon, Excel-based dispute resolution, and a carrier collaboration portal so difficult to use that carriers bypassed it in favor of emails and phone calls.
The data architecture compounded every problem. LTL rate management maintained manually in Excel by a single analyst, uploaded to IA via Transporeon. Ocean FAK rates managed on Transporeon separately. No integration between the two. When ocean surcharges updated — which happened bi-weekly — there was a lag before IA reflected the change. During that lag, every invoice processed against the stale rate was a potential error. The first-time match rate through IA was approximately 70% — meaning 30% of invoices required manual exception handling before approval. LTL alone generated 200-plus invoices per day. One analyst spent 25 hours per week managing IA’s exception output.
The specific failure modes were documented: invoices incorrectly addressed to a sister retail banner — carriers with duplicate entity records — moving through IA and into the payment queue without entity validation. Bangladesh orders — multiple BOLs and charges against a single container — producing duplicate charge scenarios IA could not resolve. Prepaid and collect term mismatches processed incorrectly. And a reporting landscape fragmented across Alteryx, Power BI, and IA outputs that required manual reconciliation before anyone could see a complete picture of what had been paid.
What Freehand Did
Freehand replaced Intelligent Audit with a 5-agent AI architecture — Invoice Ingestion Agent, Costing Agent, Audit and Claims Agent, GL Coding Agent, and Analytics and Insights Agent — each with transparent, explainable decision logic rather than IA’s black-box workflow. The Invoice Ingestion Agent accepts EDI 210/110 from high-volume carriers and email-based PDFs from the rest, normalizing every format into a single data model. Entity validation runs at ingestion: invoices addressed to the wrong retail entity are flagged before the audit even starts. Bangladesh container scenarios — multiple BOLs against a single container — are resolved through a unique identifier combining PRO number, BOL, and container number.
The Costing Agent pre-calculates expected charges for every shipment using the company’s rate cards before invoices arrive — Transporeon for ocean, Excel/contract sheets for LTL, FedEx and UPS portal rates for parcel. When an invoice arrives, it is compared against the pre-calculated expected charge. Bi-weekly ocean surcharge updates flow into the Costing Agent automatically on the Transporeon publishing schedule — no manual re-upload, no lag period during which invoices run against stale rates. The Audit and Claims Agent validates every line item, categorizes every exception, and manages carrier communication via the Collaboration Agent with structured evidence packets and tracked resolution.
The GL Coding Agent automates cost center assignment for every invoice — handling prepaid and collect term distinctions that IA had processed incorrectly, mapping inbound shipment costs to PO codes, allocating ocean charges across multi-BOL container configurations. The Analytics Agent replaces the fragmented Alteryx/Power BI/IA reporting landscape with a single intelligence platform — real-time first-pass match rate tracking, exception trend analysis, and savings captured versus leakage identified. The $800K–$1.3M annual recovery is documented, measurable, and traceable back to specific invoice discrepancies that the 5-agent system catches before payment. The company’s commitment to innovation — the retailer that invented one-stop shopping — now extends to how it manages freight spend.
