Fleetworthy Launches RADAR AI Layer Targeting Toll Leakage for Fleets

Fleetworthy has released a large language model tool called RADAR — Real-time AI, Decisioning, and Automated Resolution — in limited access, promising fleets a natural-language route to finding toll leakage and compliance bottlenecks that historically required hours inside reports. General availability of the tolling functions is not scheduled until the first quarter of 2027, a gap of roughly two years of staged rollouts between announcement and full deployment.

The safety, compliance, and toll bypass technology company unveiled the LLM functions at its third Roadshow customer conference in Nashville, giving customers an early preview before the formal launch. For now, RADAR is exactly one thing: a natural-language interface sitting on top of Fleetworthy's tolling data.

What does RADAR actually do today?

RADAR operates natively inside the existing Fleetworthy dashboard. Fleet managers interact with it through a chat menu and proactive data alert cards.

Through the chat interface, users can ask questions about their toll activity — why a charge appeared, or how a transponder is performing — and RADAR answers using the fleet's own Fleetworthy data. The tool can also surface data insights directly on the dashboard, flagging anomalous tolls, spending trends, and savings opportunities, along with recommended next steps.

"Our goal is to create a virtual assistant that scrutinizes the data, identifies optimization opportunities and recommendations, and proactively informs fleet managers to improve operations," said Levi Sorenson, VP of AI Strategy at Fleetworthy.

Pilot customers and Fleetworthy's customer advisory board shaped the feature set. Chief Product Officer Shay Demmons said the pilots delivered two concrete wins: "They've been able to quickly identify toll leakage that otherwise was difficult to find," and they found a faster way to ask basic operational questions — where transponders are, which trucks lack them — without digging through reports.

Why the data groundwork came first

Fleetworthy formed its AI team a year and a half ago and created a dedicated "AI product leader" position. Its acquisition strategy, Demmons acknowledged, was sometimes driven by AI talent as much as product: "Some of the companies that we acquired were because we knew they were diving into AI. They were leveraging different ways that we thought was unique."

That acquisition-heavy history left a mess underneath. Before any LLM could work, Fleetworthy had to clean and structure legacy data spread across multiple acquired platforms.

"The reason we can leverage AI in a solution like RADAR is because we did all the work at the data level," said President and Chief Revenue Officer Michael Precia. "Rationalizing the data, making sure that we can layer on things that are very valuable to customers was important, and we did a lot of work in the last year to do that."

Demmons framed the scope in operational terms: "What we really wanted to do is focus on safety concerns, things that might keep assets or drivers off the road, and toll leakage. That's the baseline for what RADAR is about: to bring those things up to the surface, so that the back office can focus more on 'where should we be focused to help move the economics of our fleet forward?'"

What comes next — and what it costs

The release schedule runs through 2027:

• Tolling functions reach general availability in Q1 2027.
• A closed beta for compliance and bypass technologies opens in Q1 2027.
• Agentic AI actions and more guided workflows arrive sometime in 2027.

The automated actions under development include automated toll disputes and transponder inventory reordering. "A follow-on phase to this that we're working on is taking those insights and moving them to workflows, so that we can start to solve the problem for them on the fly," Demmons said. "We don't want to stop at 'hey, here's the highest areas you should know about,' but more so, how do you actually resolve those issues."

Eventually, the system aims to give back-office teams a single operational picture. "We want to interconnect those things, where you can start to say 'Holistically, what is going to have the biggest impact on my business in terms of cost, risk, and compliance?'" Demmons said.

Access today is limited to a handful of interested customers, though Demmons said Roadshow attendees can sign up on request. Pricing remains unsettled; Demmons expects some parts of RADAR to reach all customers, while advanced features such as workflows will likely carry a premium. With firm pricing unannounced and agentic capability nearly two years out, fleets weighing RADAR today are buying a toll-focused query tool on the promise of a much broader automation engine still to come.