The Problem
The company distributes high-protein, low-sugar sports nutrition products through Walmart, Target, Kroger, Meijer, Walgreens, and CVS — approximately 500 orders per week from four US shipping locations. Its $33M annual freight spend ($18M LTL, $15M FTL) moved through two separate providers whose arrangement created a structural governance problem: Roadtex, the asset-based LTL carrier, also handled the LTL freight audit. The auditor was auditing its own work. Whether intentional or not, the incentive to find and recover errors from its own invoices was zero.
The FTL side was worse. Loadsmart managed the FTL broker relationship and carrier assignments. No formal audit existed for the $15M in annual FTL spend. Invoices arrived from carriers, were reviewed at a high level for reasonableness, and were approved. Whether charges were contractually correct, whether accessorials were entitled, whether weights matched what was actually shipped — none of these questions were answered systematically.
Two weekly reports — Ready-to-Pay and Discrepancies — surfaced 3 to 25 flagged lines each cycle. Even approved invoices required spot-checking because discount errors, refer/dry mismatches, and calculation inconsistencies appeared regularly. The reports identified the problem; they did not fix it.
Documentation gaps compounded the audit problem. 10 to 15% of invoices arrived missing BOLs or PODs. Retrieval required manual email coordination because documentation arrived in weekly ZIP files without indexing. If a POD was missing and a dispute needed to be filed, someone had to go find it — by email, from a carrier portal, from a folder full of weekly ZIP archives. Aging invoices sat in the queue for 30 or more days. The working capital impact was real. Carrier frustration was building.
The decision to replace the current model was driven by a specific realization: the conflict of interest was not fixable within the existing vendor structure. A broker that audits its own invoices cannot provide objective audit coverage by design. The only solution was an independent AI platform with no financial relationship with any carrier — one that audits every line on every invoice against the contracted rate, with shipment data corroboration, before a payment decision is made.
What Freehand Did
Freehand deployed 100% independent invoice audit coverage across both LTL and FTL for the first time — the Audit Agent replacing the broker audit on the LTL side and introducing systematic audit on the FTL side where none had existed. Every invoice is matched against the contracted rate in the Rate Manager Agent, against actual shipment data from Roadtex's daily Excel reports and Loadsmart's broker reports, and against accessorial entitlement rules configured for each carrier and load type. The conflict of interest is structurally eliminated: Freehand has no carrier relationship, no incentive to pass charges, and no reason to approve what is not contractually correct.
The Document AI Agent resolved the missing documentation problem at the root. Rather than waiting for BOLs and PODs to arrive in weekly ZIP files and then chasing missing ones by email, the agent monitors carrier submissions continuously, identifies missing documents at ingestion, and sends structured automated requests to carriers via the Collaboration Agent before the invoice enters the audit queue. The manual email coordination that had consumed hours each week is now an automated workflow with tracking, escalation logic, and a complete document audit trail.
Invoice aging moved from 30-plus days to under 3 days. The Audit Agent processes invoices on receipt — not in batch cycles that run weekly. Approved invoices generate an Ok-to-Pay notification to the company's AP team in NetSuite, where payment execution runs. Disputed invoices route immediately to the carrier with structured evidence packets assembled by the Dispute Agent. The carriers get paid faster on clean invoices; billing errors are caught and disputed before any payment is made rather than recovered after the fact.
The $495K–$825K in annual recovery represents 1.5–2.5% of the $33M freight spend — the systematic leakage that the broker audit was not catching and the FTL non-audit was not looking for. The 5-month payback period is the combined output of recovery value, FTE redeployment ($120K–$140K annually in reduced manual workload), and eliminated spend on a system that had been structurally incapable of finding the problem it was paid to find. Nicole Sigrist and Rosalia Zamudio, who previously reviewed every discrepancy report manually, now review only the exceptions the Audit Agent could not auto-resolve — which, at steady state, is a fraction of the prior weekly volume.
