The Problem
The company processes over 100,000 orders daily and touches one billion lives worldwide through pharmaceuticals, medical devices, and consumer health products across 400 globally managed sites. Its freight audit function had been outsourced to CT Logistics for 20 years — a relationship that had served its purpose in an era before AI-native audit was possible, but had calcified into a structural dependency. The audit was not real-time. The data was not self-service. When the logistics finance team needed to understand what had been paid, to which carrier, at what rate, for which shipment, the answer required engaging CT Logistics, waiting for outputs, and manually reconciling what came back.
The 15-global-ERP landscape was the foundational data problem. Fragmented source data across SAP and other systems was not updated in real time, resulting in repeat data entries, invoice discrepancies, and manual processes that consumed significant team hours identifying carrier billing errors across disparate systems. Rate cards for TL, LTL, Air, FCL, and LCL modes were managed with too much reliance on tribal knowledge across regions, teams, and business units. Static freight cost calculations increased the risk of overpayments that accumulated without systematic detection.
The competitive evaluation was a formal RFP — the company ran a structured procurement process against established FAP vendors. The requirement was not incremental improvement on the CT Logistics model. It was a replacement of the entire operating model: direct ownership of the audit process, AI-driven anomaly detection at line level, and a real-time intelligence layer that connected freight spend to the financial systems where it ultimately needed to land. CT Logistics had been paid to process. The company now needed a system that audited.
What Freehand Did
Freehand replaced CT Logistics across the full North American freight network — displacing a 20-year BPO relationship with an AI platform the company owns and operates directly. The Audit Agent runs multi-way matching on every invoice across Air, TL, LTL, FCL, and LCL modes: invoice against contracted rate, rate against shipment actuals from the carrier, shipment actuals against the GL coding rules that determine how cost lands in each of the 15 ERP systems. Anomaly detection identifies overcharges, duplicates, and uncontracted accessorials at line level before any payment is released.
The Knowledge Graph Agent normalized the fragmented source data across 15 global ERPs into a unified spend intelligence layer — every carrier record, every rate card, every shipment reference mapped into a consistent data model that the audit system can query in real time. The tribal knowledge that had lived in regional teams and in CT Logistics’s systems is now a managed, versioned data asset that the company owns. When a rate changes in one ERP, the Rate Manager Agent reflects it automatically. When an invoice references a shipment that does not match any source system record, the Audit Agent flags it before payment.
The 6% combined savings — 4% in primary movement through AI-powered load optimization across 400 globally managed sites, and 2% in secondary movement through intelligent rate structuring and hybrid PTL/FTL transportation models — is the financial output of replacing a process designed to approve payments with one designed to audit them. The invoice cycle time reduction through complete digitization of freight audit processes across 100,000-plus daily orders freed the logistics finance team from the manual validation work that had consumed their capacity under the BPO model. They now focus on strategic oversight and carrier management rather than on chasing discrepancies in CT Logistics’s outputs.
