“Where There’s Mystery, There’s Margin.” Flexport Wants to Fix That
The company’s new trade model pushes AI into the physical flow of goods
“Where there’s mystery, there’s margin,” said Flexport CEO Ryan Petersen during the company’s 2026 Winter Technology Release.
He was describing global logistics. An industry where limited transparency into routing, customs filings, and carrier performance often translates into higher costs.
Flexports says its latest release is built around reducing that mystery. Flexport released its Winter Technology update, outlining a set of AI-driven supply chain tools including a public trade data model called Atlas. The release centers on automation, compliance, and execution infrastructure rather than standalone analytics.
Flexport is arguing that it is not simply automating freight workflows. It is building a shared data model of global trade and tying that model directly to freight execution and customs compliance.
Atlas as Trade Infrastructure
At the center of the release is Atlas, an interactive model of the global ocean freight network. Flexport describes Atlas as exposing port congestion, sailing schedules, service strings, and vessel data that mirror the dataset used internally by its operators and algorithms.
The company also says Atlas data is accessible through the Model Context Protocol (MCP), an open standard designed to connect AI systems to structured data sources. Anthropic’s technical documentation describes MCP as a way to allow AI agents to securely access external data and tools.
Flexport’s stated objective is to use a single structured model to power human operators, planning algorithms, and future AI agents.
That model feeds several execution layers.
Digital routing guides replace the spreadsheets many large shippers use to allocate carriers. Instead of static contract hierarchies, routing logic can be encoded (lowest cost, best reliability, or specific allocation ratios) and translated directly into bookings inside Flexport’s platform.
The release also introduces AI-powered search within Flexport’s Control Tower and reports an 80 percent reduction in notification volume through relevance filtering and user-specific routing logic.
Flexport further reports that its AI-driven shipment consolidation engine has produced roughly 10 percent freight cost reductions for participating customers. Independent industry case studies show that freight consolidation can generate double-digit savings under certain shipment conditions.
Other vendors emphasize visibility. project44 markets real-time tracking and predictive ETA capabilities across global carriers through its Movement platform. FourKites promotes an “Intelligent Control Tower” designed to provide predictive analytics and multimodal freight monitoring.
These systems aggregate and analyze carrier data. Flexport’s distinction is that its trade model feeds routing, bookings, and customs entries within its own execution stack.
The Winter Release places heavy emphasis on customs compliance.
Flexport states that its AI audit system reduced its internal customs filing error rate from 1.8 percent to 0.2 percent by auditing 100 percent of entries in real time. It also launched a public broker audit tool that allows importers to review customs filings using government ACE data.
The company reports $900 million in cumulative customs savings and refunds delivered to customers over the past five years, including duty drawback recoveries identified through algorithmic matching.
Regulatory authorities such as the World Customs Organization publish compliance guidance for brokers but do not provide comparative public error-rate benchmarks.
These compliance tools were introduced as tariff enforcement rules shift following the Supreme Court ruling. While CBP halted collection of the invalidated tariffs, Reuters reported that refund processes remain uncertain.
Flexport’s Tariff Refund Calculator is positioned as a preparatory tool for estimating potential claims tied to those duties.
The competitive distinction rests on integration.
Visibility platforms do not execute freight. Traditional forwarders execute freight but historically operated across fragmented systems. For example, companies such as C.H. Robinson market integrated digital freight tools layered onto their execution networks through platforms like Navisphere.
Flexport’s strategy is to unify those layers under a single data model.
Atlas feeds routing logic. Routing feeds bookings. Bookings feed customs filings. Customs filings feed audit and refund tools. The same structured data underlies each step.
Rather than positioning AI as a separate analytics layer, Flexport is embedding it into freight execution and regulatory compliance.
As tariff rules shift and enforcement uncertainty increases, the company is arguing that control over both data and execution matters.