Most transportation management systems were built to answer a simple question: What’s the cheapest way to move this order? What they’ve been slower to answer is whether that plan will actually work once a driver, a dock door and a delivery appointment enter the picture. Then add in multiple dock doors, several delivery appointments, hours of service regulations and more, and there is a lot of complexity for the average planner to work through.
That gap is where FreightWise is placing its bet with Kuebix AI, an AI-backed TMS for small and mid-market shippers. Announced Oct. 20, the platform becomes the single system for every FreightWise and Kuebix customer, covering LTL, truckload and parcel in one place.
FreightWise CEO Chris Cochran sat down with FreightWaves at the F3: Future of Freight Festival in Chattanooga, a day after demoing the platform to attendees. His pitch was direct about who the product is for. “We’re focused 100% on the shipper,” he said, contrasting Kuebix AI with the many technology providers at the event that build for brokers.
One platform, one roadmap
The launch is partly a consolidation story. FreightWise acquired Kuebix in 2023 and has since operated two separate systems: FreightWise TMS and the Kuebix TMS platform. According to the company’s announcement, every customer will now work on the same technology, supported by one team and guided by one product roadmap. The practical upside, FreightWise says, is that features built in Kuebix AI reach every user faster.
Cochran described the starting point as a legacy platform with a loyal user base. FreightWise already ran a modern, native application for its managed transportation business, which gave the team a head start. They brought the features Kuebix users knew into that platform, and the result became Kuebix AI.
“Over the past year, our team has been rebuilding Kuebix with the same functionality Kuebix users have known for almost two decades, while innovating on the infrastructure, aligning AI with core features, and ensuring the platform has real staying power,” Cochran said in the announcement.
Thousands of users are already using Kuebix AI, and upgrades onto Kuebix AI will continue for SaaS only customers in the coming year.
Replacing the black box on multi-stop routes
The feature Cochran emphasized most is multi-stop truckload planning, which is where he sees the biggest gap between how TMS platforms work today and how planners need them to work.
In a traditional system, he explained, the multi-stop route is planned and then handed to a planner, who often discovers that it will miss delivery appointments. The planner can’t see why the system built the route that way and usually has no alternatives to choose from.
Cochran said he sees that as a legacy of optimization logic written years ago. “AI disrupts that in general because AI doesn’t need those algorithms,” he said. “It’s smart on its own.”
In Kuebix AI, the planner sees several route options on one screen, along with where each one is weak. One option might add miles to hit a delivery appointment. Another might cut miles to save cost but force a change to the appointment. The planner drags over the one that works and acts from there, whether that means resetting a customer’s expectations or flagging a retailer shipment where fees or damages could follow.
The point, Cochran said, is to keep planners inside the TMS. “Instead of them going away from the TMS when they have a bad route or a false positive, it allows them to do it right there.”
The underlying pieces highlight AI-driven service time builds delivery plans around hours of service, wait time, loading and unloading, and facility hours. Multi-stop AI routing then uses that data to flag whether a sequence of stops is feasible and recommends the optimal route. FreightWise says the goal is for the delivery dates shippers give their customers to hold up.
Integration without the IT queue
Integration is the other pain point Kuebix AI solves. A TMS is rarely the system of record for orders, Cochran noted, and in the SMB and small enterprise segments, IT resources are scarce and the systems feeding data are often old. Few people can work on them, and they may not support current JSON calls.
Kuebix Connect, a no-code middleware layer, is FreightWise’s answer. Cochran said the aim is to put control in the hands of the business user. That person matches fields between a WMS, ERP or other source system and the TMS visually, and maps the logic. Then AI within Kuebix Connect will monitor the connection on an ongoing basis.
“As the daily feeds come through, sometimes there might be bad data, or failures in feeds,” Cochran said, usually from something small like an incorrect character that keeps orders from coming into the system. “The AI embedded within Kuebix Connect will monitor that, will fix it and help you diagnose the source of that” so that freight doesn’t sit in the warehouse waiting on a data fix.
IT still has a role, he added. They’ll need to provide credentials and test for security, but the company isn’t relying on IT to make manual calls every time something breaks.
Dashboards that explain themselves
The third piece is Kuebix Intelligence, which the company describes as AI-backed dashboards for spend by location, carrier and mode that explain the drivers behind the trends, not just the numbers. Cochran tied this to end-to-end design not only of Kuebix AI, but FreightWise’s solutions as whole. Because rating, routing, freight audit and payment all live in one system, rather than having settlement data sit with a separate audit provider, he argued that the data feeding the AI is better.
“If your KPI is off, the AI tells you why it’s off, where did it go wrong,” he said, rather than a CFO waiting on a logistics analyst for an explanation.
The release adds a related capability. AI-powered freight audit exception resolution analyzes quoted rates against final invoiced amounts, identifies discrepancies, quantifies the financial impact and recommends carrier dispute strategies.
Why FreightWise waited
Cochran was candid that FreightWise wasn’t first to the AI conversation, and that it was a deliberate choice. By his own estimate, the company moved six to eight months later than he would have liked. But the management team was wary of a trend they saw across the industry. “We were kind of uncomfortable with this trend that people were putting AI for the sake of AI in applications,” he said, pointing to chatbots running on incomplete data and no context. For some of the early adopters, users would engage, but they were not longstanding value-driven AI solutions that solved real problems.
“We waited until we could make it operational,” Cochran said.
That philosophy shows up in the company’s term for its approach: operationally executable AI, meaning AI that solves real pain points. FreightWise CTO Chris Haarmeyer echoed the point in the announcement: “We didn’t set out to add AI for the sake of it.” The team started with problems customers and internal users raised, like delivery dates that ignore hours of service or scheduled appointments.
The company also addressed data security up front. “Customer data stays protected, it is never used to train third-party models, and every AI feature is opt-in,” Haarmeyer said.
Where Cochran draws the line
Cochran’s view of AI in logistics is enthusiastic but bounded. He named two concerns. The first is data quality. “You’re not ready for AI if you don’t have good data,” he said, warning that AI can hallucinate and reach false conclusions when it’s fed bad inputs. Shippers without a clean, single source of record, he argued, will be at a disadvantage to peers who have one.
The second is customer relationships. “AI is not here to manage your customer relationships,” Cochran said. He doesn’t want customers calling FreightWise support or an account manager and ending up talking to a bot. When a customer is in a pinch, he wants them talking to a real person.
He also isn’t positioning AI as a cure-all. Someone still has to sell the order, load the truck and deliver the goods, he noted. But he was firm that shippers who skip it risk watching their cost to serve drift out of control.
What to watch
For shippers, the proof will come in execution. The release says additional AI-backed functionality will roll out in 2027 and beyond, and the upgrade of legacy Kuebix users to Kuebix AI is actively underway. The priority continues to be ensuring new features are solving real business problems, easy integrating for shippers with their existing technology with AI to flag any errors, and prioritizing data to inform better decisions.
The post Inside Kuebix AI: A Shipper TMS Rebuilt Around Executable AI appeared first on FreightWaves.
