Tech / 'AI won’t replace forwarders – but forwarders using AI will'

© Anatolii Savitskii |

Freight tech firms love to call themselves disruptors. But real transformation in forwarding isn’t about flashy tools; it’s about reimagining the underlying processes that have shaped the industry for decades, says Raft’s CEO and co-founder James Coombes.

“We are absolutely a tech company, an AI company,” he says. “But we are also a change-management company.”

Technology alone doesn’t deliver value, unless organisations are willing to challenge long-held assumptions. And in freight forwarding, where any given workflow can span continents, time zones, and dozens of partners, that requires honesty about how things actually get done.

Mr Coombes says the biggest mistake companies make is trying to “plug a gap in an existing process with AI”, but leaving the overall workflow the same. That produces, he says, the classic square-peg-round-hole problem; an automated process running inside an unchanged higher-level system, with unexpected, and sometimes counterproductive, knock-on effects.

Instead, he encourages customers to imagine what the whole process could look like when automation is baked in from the start.

“We don’t come in with a ‘hey, we know best’ attitude” he says. “We help customers imagine a world where technology is a foundational part of the solution, and work from there to reinvent processes that may be deeply established.”

James Coombes,        Raft CEO and founder    

Forwarders often think of AI as a tool for extracting data “more effectively, with fewer errors”. But Mr Coombes says that’s only the beginning. The more interesting question is what becomes possible after the data is standardised.

In forwarding operations, especially customs, people often extract only the minimum necessary information from documents. “But if you were to do an extra 5% or 10% and get more data in a more standardised way, that would benefit downstream applications,” he explains. A small lift at one step can unlock huge efficiencies elsewhere, particularly with multiple teams.

That’s why the old saying, ‘you can’t manage what you can’t measure’, is painfully true in forwarding.

“Existing data is all over the place,” Mr Coombes notes, “because it’s human nature. We cut corners. Bob’s data standards are different from Susan’s.” AI can fix that inconsistency, and give management the visibility it’s lacked for decades: why one process or customer or tradelane underperforms; where rate decisions go wrong; where capacity breaks.

But is it also a threat? Mr Coombes is dismissive of the trope that AI will replace staff. “Fear of AI is overblown,” he says. “We’re not here to replace an individual, but make them more effective and more efficient. AI won’t replace forwarders, but forwarders using AI will.”

Operators get time back; less manual data work and fewer repetitive tasks, while managers get something even more valuable: “a supercharged team” and clear, actionable data.

This shift also allows forwarders to escape the transactional nature of their business. “The value is in the relationship with your customer,” Mr Coombes says. “If you take away the automatic, repetitive work, it allows forwarders to be the true problem-solvers they are.”

If freight forwarding has a constant, it’s unpredictability. Volumes spike and collapse. Surcharges appear overnight. Capacity shifts across modes and geographies. That’s where scalable automation becomes essential, he argues.

“The number-one view is: how do we allow our customers to scale more effectively with technology?” Mr Coombes says. Raft users routinely report being able to handle “two or three times more shipments per person” once the system is embedded.

This scalability has become strategic. Many forwarders now find their customers are also driving change, asking: ‘What’s your AI strategy? How prepared are you for the future?’ And they expect an answer.

Mr Coombes is quick to draw a line between consumer AI and enterprise-grade AI. In a personal setting, a hallucinated answer is an annoyance. In freight forwarding, it can have commercial, legal, or safety consequences.

“Everything has to be auditable, accessible by humans, collaborative,” he stresses. Enterprise AI must come with guardrails, quality controls, and transparency, he explains.

In his view, the future is clear: every forwarder will use AI because competitiveness demands it. Margin pressure isn’t going away. Time-zone constraints aren’t going away. The need to do more with less isn’t going away.

“Automation helps answer things when you’re sleeping,” he says simply.

And in a global, always-on supply chain, that may be the most compelling argument of all.