FourKites launches AI model to deliver supply chain outcomes

FourKites has launched an Outcome-as-a-Service model for supply chain customers and aims to use AI agents to complete one billion hours of operational work for them over the next decade.

The shift aligns FourKites' commercial model with the business targets customers are already judged on, such as on-time-in-full performance, expedited freight spend and cost to serve, rather than adoption of a particular software tool.

Under the model, engagements begin with a specific operational metric and work backwards to the data, workflows and AI agents needed to improve it. FourKites said the approach reflects a broader shift in enterprise software buying, with customers placing greater weight on completed work and measurable results.

Mathew Elenjickal, Founder and Chief Executive Officer of FourKites, said the company was moving beyond the visibility tools on which it built its business.

"We founded real-time visibility to show customers where the problem was. Outcome-as-a-Service is how we solve it for them. Every leader we work with is accountable for an outcome, not a capability. A VP of supply chain isn't measured on whether they've adopted AI. They're measured on whether product arrives on time and in full, at the right cost," said Mathew Elenjickal, Founder and Chief Executive Officer of FourKites.

Network data

The model relies on FourKites' broader supply chain data network, which the company argues gives its AI agents external context that a single company's internal systems cannot provide. That includes information such as carrier performance on a route and conditions at a port or receiving dock.

Its network graph connects more than 1,600 brands, more than 500,000 carriers and 3.6 million facilities across 176 countries. FourKites said those data points are used by its AI agents before each action and fed back into the system after each result, with the aim of improving decision-making over time.

The emphasis on external operational context comes as many large companies continue to run supply chains through multiple systems. FourKites said a typical Fortune 2000 supply chain uses 10 to 15 separate tools, each optimising a single function, with few linking directly to broader cost, service and revenue objectives.

Elenjickal said access to real-world network data was central to the company's position in the market.

"AI models are becoming a commodity. What can't be copied is real-world network data and the context around it. That context is what lets our agents act on a customer's behalf, and it's why we can commit to outcomes," said Elenjickal.

Internal overhaul

Over the past year, FourKites has also changed its own operations to support the new model. The company said it rebuilt internal functions to operate in an AI-native way, including automating its software development lifecycle end to end.

As part of that change, the business moved from releasing software every two weeks to making releases daily. FourKites also replaced its traditional product catalogue with a digital structure organised around more than 70 outcomes, each linked to the AI agents used to execute them.

The new approach points to a shift in how software suppliers package and sell AI tools to large companies. Rather than leading with technical features, FourKites is framing its offer around the operational work performed and the business measures affected.

That may resonate with supply chain executives under pressure to improve service levels and manage costs while dealing with fragmented technology estates. It also puts pressure on FourKites to show that autonomous systems can move from alerting staff to taking action on their behalf with consistent reliability.

For FourKites, the announcement marks a broader repositioning of the business more than a decade after it helped establish real-time transportation visibility as a distinct category in supply chain software. Its mission is now defined not only by building the network behind those decisions, but also by measuring the amount of work completed for customers through AI agents.