AI agents are set to become commoditized within months, and in supply chain the durable competitive advantage lies in the data foundation underneath them, according to Valentina Jordan, CEO and co-founder of Nauta. In an interview with FreightWaves, Jordan explained why AI for global trade only works when the movement of goods, data, and money are connected, describing Nauta's "operational brain," how it learns from ERP, TMS, WMS, email, and spreadsheets, and why messy data remains the biggest blocker to real AI returns. The conversation also covered Nauta's strategic funding round, its global expansion, and how AI can help operators cut stockouts, detention, and demurrage by turning fragmented workflows into action.

Nauta, an AI-native operating system for global trade, has closed a strategic funding round backed by BMW i Ventures, Bosch Ventures, Hitachi Ventures, and Yamaha Motor Ventures, Jordan announced. The company is deploying the capital toward global expansion; it already operates in more than eight countries with 8,000 suppliers and reach across 60 countries through its clients.

Jordan was direct about her competitive thesis in a crowded AI-for-supply-chain market: the agents themselves are not the differentiator. "We think that agents are going to be commoditized — I'm not going to say weeks, but probably months," she said. "The real challenge here is the data underneath those agents, and especially in supply chain and global trade, that data needs to be all about the context, knowledge, and experience that operators have built over decades."

"We do not start agent first. We start data first, output first, and then comes the agent."

The argument tracks a broader shift in enterprise software: as large language models and agent frameworks become widely available from major cloud and AI vendors, differentiation has migrated from the models themselves to proprietary data, workflows, and domain context. Supply chain has proven a particularly hard case for that thesis, because operational knowledge is often trapped in emails, spreadsheets, and the heads of veteran planners rather than in structured systems.

Nauta's platform — what Jordan calls an "operational brain" — connects to a company's existing ERP, TMS, WMS, and OMS systems, as well as unstructured sources such as email, Microsoft Teams, and spreadsheets, where Jordan says operators conduct 80 to 90% of their work. The brain continuously ingests contracts, arrival notices, payment terms, and external signals ranging from weather events to port disruptions, building a layered model of how each individual company operates. Jordan cited a recent earthquake in Colombia that affected local ports as an example of the macro-level signals the system monitors.

On top of that data infrastructure, Nauta deploys task-specific agents. One high-value use case is three- and four-way invoice matching for accounts payable reconciliation; Jordan said the system proactively identifies between $300,000 and $500,000 in incorrectly invoiced amounts per client. Other active use cases include overage, shortage, and damage claim building; freight audit; and inventory management focused on reducing stockouts and overstock to improve cash-to-cash cycles. These are long-standing pain points in freight operations, where audit and claims processes have historically depended on manual reconciliation across shipping documents, invoices, and purchase orders.

The investor lineup — all large industrial and automotive conglomerates with global footprints — reflects what Jordan describes as a shared recognition that fragmented data is blocking AI adoption at scale. Customers span food and retail distribution, manufacturing, and consumer brands including New Balance, L'Oréal, and Moët Chandon, and Jordan said the data problem is consistent across all of those verticals.

"There's more similarities between a food distributor and a furniture or chemical business than there is between two food distributors, because every company operates differently," she said.

Jordan traces Nauta's operational credibility to her co-founder, Rafa, a third-generation Spanish immigrant to Puerto Rico whose family operated a 123-year-old food distribution business. That firsthand experience with logistics pain points shaped the company's conviction that visibility platforms alone cannot solve the underlying data problem. Jordan herself comes from a tech background spanning 13 years, including stints at Amazon and Rappi, a last-mile delivery company.

She framed the funding round as both a validation of Nauta's data-first thesis and a launchpad for international growth, arguing that without a clean, contextualized data foundation, large global enterprises will continue to struggle to extract measurable value from AI investments. How quickly agent commoditization actually plays out — and whether data-layer startups can defend their position as incumbent ERP vendors build their own AI capabilities — is likely to shape the next phase of competition in AI for supply chain.

Source: FreightWaves