Flexport reports 40% freight volume growth

Flexport has reported a 40% year-on-year increase in freight volumes alongside a 5% rise in its workforce, with founder and chief executive Ryan Petersen identifying artificial intelligence as a contributor to the company’s operating performance. The growth accompanies the wider use of automated customs processes, transport planning technology, and software tools across its international freight forwarding operations.

Petersen outlined the figures during an interview published on 2 November, following Flexport’s autumn technology release. The company has been expanding the functions available through its digital platform, including customs classification, automated business rules, and multimodal routing, while developing AI tools for internal operational and software engineering work.

The reported volume and staffing changes indicate that Flexport is handling more freight without an equivalent increase in employee numbers. Its shipment mix, operating conditions, and customer demand also influence that relationship, while the company attributes part of the difference to the use of automation across its processes.

Customs brokerage provides one of the clearest examples of how the technology is being applied. Imported products must be classified using the appropriate Harmonized Tariff Schedule codes, with the classification determining the applicable tariff treatment and supporting the information required for regulatory declarations.

Classification can require detailed examination of a product’s construction, materials, function, and intended use. Different goods may appear similar in commercial descriptions while falling under different tariff provisions, making accurate technical information and the interpretation of customs rules essential to the process.

Flexport has introduced an AI-assisted classification tool that guides users through questions about a product and recommends a ten-digit United States tariff code. The system presents supporting reasoning and links to relevant US Customs and Border Protection rulings, allowing users to compare the suggested classification with applicable source material.

A licensed customs broker reviews the proposed classification before it is used in a customs filing. The arrangement allows software to assist with researching possible codes while keeping the regulated classification and declaration process under the control of an authorised professional.

Historical customs decisions create an additional challenge because published rulings may remain accessible after subsequent decisions change their applicability. A classification recommendation therefore requires consideration of the current rules and the specific characteristics of the goods being imported.

Flexport’s compliance technology also supports customers whose shipments must follow established internal procedures. The company has developed a business rules engine that uses AI to help translate customer instructions into executable software rules, which can then be reviewed and applied consistently to relevant customs entries.

A business might require particular documentation, checks, or handling procedures before an import declaration is submitted. Converting those requirements into repeatable controls allows the system to examine transactions against the agreed instructions and direct exceptions towards the employees responsible for resolving them.

Once the rules have been tested and approved, their routine execution can be automated. The process reduces the requirement for staff to interpret the same instruction repeatedly, while changes in customer requirements or customs regulations still need to be incorporated into the approved procedures.

Flexport has also developed tools for auditing customs records and identifying potential errors or opportunities to correct previously submitted entries. Petersen reported that the company found substantially higher error rates in competitor entries examined through its own audit process than in Flexport’s US customs filings.

The results form part of Flexport’s assessment of its customs operations and support its investment in automated checking. Individual audit findings still depend on the records examined, the rules applied, and the review procedures used to determine whether a potential error requires correction.

Transport planning has been expanded through Flexport Atlas, which originally concentrated on ocean freight routes. The platform now incorporates air, rail, and barge connections, allowing users to examine transport movements that involve several modes rather than viewing maritime services in isolation.

An international shipment may involve a feeder vessel, a mainline container service, and an inland rail journey before reaching its destination. Incorporating these stages into a common route view allows users to examine the sequence of connections, the expected time required for each movement, and the availability of alternative routes.

The Atlas platform includes historical transit information, predicted dates, disruption monitoring, and emissions estimates for specified transport arrangements. The information can support route selection and shipment planning, although the final service available to a customer will depend on carrier schedules, capacity, and the terms of the freight booking.

Flexport has also introduced a Model Context Protocol server allowing authorised AI systems to interact with parts of its logistics platform. The supported functions include requesting shipment information,

obtaining freight quotations, identifying exceptions, and initiating bookings using the relevant account and service arrangements.

These tools extend automation into the interaction between logistics customers and the freight forwarder. A user can obtain information or prepare a transport action through a connected interface, while the underlying booking still requires valid shipment details, suitable capacity, and the permissions applicable to the account.

Petersen has identified internal software development as another area in which AI is changing operations. Flexport is examining smaller engineering teams supported by automation, with software tools used to accelerate development work and reduce the manual effort associated with certain tasks.

Freight demand has also contributed to the company’s reported growth. Petersen said ocean freight volumes remained elevated beyond the period normally associated with seasonal import peaks, while shipments connected with data centre construction and technology infrastructure were contributing to air cargo demand.

Semiconductor components, servers, and other high-value infrastructure can create transport requirements in which shipment value and delivery timing affect the choice of freight mode. Expensive equipment held in transit represents capital committed before the goods can be installed or used, while project schedules can place additional pressure on delivery arrangements.

Petersen estimates that data centre-related cargo currently accounts for approximately 8% of global air freight capacity and expects the proportion to increase. The estimate reflects his view of market conditions and has been identified separately from Flexport’s own freight volume and staffing figures.

The company’s operational development now combines higher reported shipment volumes with expanded software capabilities across customs and transport planning. Continued growth will require the systems and specialist teams handling bookings, declarations, and freight exceptions to accommodate more transactions while maintaining the controls required by customers and regulators.

Flexport’s reported 40% volume increase and 5% workforce growth establish its current operating position. Its next technology releases are expected to extend automated logistics functions further, building on the customs, routing, and platform integrations already introduced during 2026.