The multi-trillion-dollar logistics industry runs on trucks and trains, ships and planes. It also runs on calls, texts, and email.
“We get hundreds of thousands of emails saying, ‘Hey, I have this freight that needs to move from X to Y—how much would it cost?’” says Mike Neill, chief technology officer at logistics provider C.H. Robinson.
The details vary from load to load and mode to mode, but generally a company that wants to ship a load of goods reaches out to freight brokers, who in turn connect with carriers who actually haul the goods from place to place. Getting the details right, like whether the shipper has a full truckload, specific insurance requirements, or cargo needing special handling, and when and where the carrier must send a truck, can mean a lot of back and forth that’s traditionally been handled by people sitting at phones or computers.
Now, it’s increasingly being handled by artificial intelligence, which can help all parties in the equation respond to emails, communicate availability, and even solicit bids and issue quotes, often much faster than humans feasibly can.
“We determined that the humans weren’t getting to all the opportunities,” says Neill. “They weren’t getting to it in a timely manner.”
As a result, C.H. Robinson gradually rolled out a suite of AI agents, starting with bots that simply detected whether incoming emails were asking for shipping quotes. Over time, those advanced to pull out more information from quote requests and ultimately respond to many requests automatically. That means many shippers can now get a response within seconds, helping the company win more business, increase the number of shipments processed by each employee, and boost its margins per transaction.
C.H. Robinson, considered the largest freight broker in North America in managing 37 million shipments per year, rolled out much of its AI infrastructure in-house. And other logistics companies have turned to tech companies focused specifically on applying AI to the sprawling sector, where they say it can help cut costs and boost efficiencies. That, they say, lets humans focus on the most complex problems and building relationships key to repeat business rather than spending their days playing phone tag and scurrying to answer emails.
While much of the public’s attention on transportation automation has focused on self-driving trucks and drone deliveries, AI has quietly made inroads in the offices (and semi-truck cabs) where deals are done and plans are made.
Harish Abbott, cofounder and CEO of supply chain AI company Augment, describes the pre-AI system of calls and emails as imposing a “coordination tax” that leads to waste and unnecessary delays in transporting goods. “Trucks have to wait, they go empty miles, warehouses have to have underutilized labor, or businesses have to carry more inventory,” he says.
Augment’s AI system, called Augie, can sift through brokers’ incoming email to spot shippers looking for a quote or the status of a delivery or detect when a carrier is looking to get paid or seeking more cargo to transport. The AI can generate quotes based on rules provided to Augie or by integration with an existing quoting engine, understanding when more information is needed and routing requests to a human if necessary. The system can also run automated security checks, making sure that a potential counterparty is actually someone a business wants to deal with. Since the AI logs its work, it also makes it easier to track win rates and identify patterns in winning and losing bids, Abbott says. Augie can even place calls or emails to carriers, speaking in a variety of languages as needed, and extract data from emailed updates in a variety of formats.
The company also offers its technology to carriers themselves, where it can also respond to incoming quote requests, pull important data about shipments, and communicate with drivers about their status. In some cases, Augie may even be participating on both sides of a conversation, though Abbott emphasizes that each instance of the software is acting on behalf of the appropriate client’s best interests.
“Literally every month, the number of incidences where AI is talking to AI, Augie is talking to Augie, is going up,” he says.
But Abbott and others in the industry say AI is unlikely to completely replace humans any time soon. Instead, says Robert Nathan, founder and CEO of logistics AI company Envoy AI, the technology can help shift people away from busywork like answering emails, tracking shipments, and transcribing updates from carrier to shipper.
“These are college-educated individuals doing low-level tasks, and I think that this next generation of operators deserves better,” says Nathan.
Envoy’s AI, known as Ellie, can help carrier reps, the people at freight brokers tasked with connecting with shipping carriers, negotiate pricing, communicating with carriers via email, phone, or text as appropriate. “You really try to communicate and meet the trucker or the dispatcher where they are,” says Nathan.
Like other AI providers in the industry, Envoy’s system can be configured to negotiate within existing guardrails and get human pre-approval before deals are done to minimize the effects of AI hallucinations, so businesses shouldn’t have to worry about AI proposing or accepting money-losing bids. And as Nathan points out, the AI can answer emails around the clock and doesn’t experience hunger or fatigue, factors that can impact humans in their negotiations.
In turn, human employees can spend more time building relationships with business associates, which can help establish longer term arrangements for regularly recurring shipments or make it easier to address any problems so they don’t happen again.
“Spending less time on this very tactical stuff, but now spending more time on saying, how do I strategically build a repeat business,” says Abbott. “That’s where the relationships matter, and that’s where the real majority of the business is.”
While logistics is for good reason often called a “people business,” it’s also one in which numbers, speed, accuracy, and volume matter, says Simson Demmer, chief customer officer at Cargo.one. The company provides AI focused on air and sea transport for both carriers and freight forwarders, capable of generating quotes based on its customers’ parameters. Those can include everything from bidding policies to email tone of voice. Cargo.one also has an extensive database of air and sea transport rates. And though it’s only been in recent times that AI models have been powerful enough for automated pricing and communication, Demmer says the integrations with customer records, which are also used to test the AI against historic quotes before launch, and with its own trove of industry data are at least as vital.
“If you don’t have the price for tomorrow’s flight or tomorrow’s ocean line crossing, the best AI doesn’t close that gap for you,” he says.
Carriers, too, can benefit from AI communication tools, especially since many are mom-and-pop trucking operations where dispatchers or drivers themselves must regularly reach out to brokers seeking loads to transport.
“There are hundreds of thousands of small trucking companies across the United States,” says Paul Singer, cofounder and CEO of FleetWorks. “Most of them are getting most of their freight through a phone call, an email, a text message.”
FleetWorks’ AI can proactively call brokers on carriers’ behalf and help brokers find vetted carriers able to transport goods. The Y Combinator-backed company’s network of carriers and brokers is growing rapidly, now including 40 of the top 100 freight brokers and nearly 29,000 carriers, Singer says. It’s already helped brokers’ carrier reps work significantly more efficiently. One customer recently reported a single carrier rep moved from booking roughly 150 loads per month to more than 300, he says.
“We doubled the productivity of this carrier rep by bringing that carrier rep a lot more options, and handling most of the communication for them,” he says. “They’re also spending a lot more time building relationships, because they’re not handling transactional bookings.”
Exactly how much AI will ultimately transform the transportation world remains to be determined, though companies in the industry are already pointing to those kinds of visible gains from automated price quoting and negotiation.
Logistics companies are also increasingly using AI for parsing and generating the often-voluminous sets of documents related to shipments, as well as tracking the locations of goods, handling invoices and payments, and helping carriers resolve any issues along the way.
To AI enthusiasts in the industry, the sheer volume of verbal communication and written documents—which until recent years were seen as an impediment to automation—are a clear sign the field is ripe for efficiency gains from large-language models.
“There’s just so much unstructured data and so much opportunity within these businesses,” Singer says, “because all this data has been sitting in siloed systems of record for a very long time.”