A truck breaks down. Someone calls a shop. The shop says they'll get back to you. Two days later, nobody has. This is still how repair works for most fleets in 2026, and it's been this way for about as long as anyone can remember.

Today ServiceUp is launching the next generation of its platform: the agentic repair platform for modern fleets, built around AI Repair Agents that handle the parts of this process a person used to have to push forward by hand.

Why this needed to change

Ask a fleet manager what fills their day and the answer is rarely "managing the fleet." It's a vehicle sitting for three days because nobody followed up on an estimate. It's stepping away from the job to get a shop on the phone. It's another spreadsheet, another status check, another email that says nothing new.

None of that shows up on an invoice. All of it shows up in the numbers that matter more, cost per vehicle, time in service, how much a team can get done in a week.

"Fleet repair has run the same way for decades. Phone calls, emails, a fleet manager stuck playing middleman," said Brett Carlson, CEO of ServiceUp. "Nobody questioned it, because nobody thought there was another way to do it. We did."

What's launching

ServiceUp sits between a fleet's existing system of record and its shops, and takes over everything that happens in between. A repair can start anywhere, an FMS, a telematics alert, a driver calling it in, or a shop where the vehicle just shows up unannounced, and it gets turned into a structured repair order automatically, regardless of where it started.

From there, the agents take over the parts of the job that used to eat a fleet manager's week. One finds the right shop for a repair and routes it there without anyone picking up a phone. Another checks every line of an estimate against a fleet's pricing, warranty terms, and maintenance policy, and clears it in seconds instead of days. A third watches every open repair around the clock, so a slipping deadline or a cost that looks off gets caught before it turns into a bigger problem instead of after. And everything, every photo, every message, every approval, lives in one place tied to the job itself, instead of scattered across someone's inbox.

Fleets stay in control of how far any of this goes. Run it fully manual and agents just surface recommendations for a person to approve. Run it in co-pilot and agents act on their own up to limits the fleet sets, with anything past that going to a queue for review. Run it in autopilot and agents handle the whole process end to end, stepping back only when something genuinely needs a person's judgment.

"We don't think downtime and paperwork are just the cost of running a fleet," Carlson said. "They're the cost of running one manually. That's the part we set out to fix."

What changes for the people doing the work

The clearest way to measure this isn't a percentage, it's a Tuesday. It's a fleet manager who used to spend the first hour of every day on the phone with shops, now spending that hour on the parts of the job that needed them most. It's an estimate that used to sit in an inbox for two days getting approved before the shop has even finished the inspection. It's a vehicle that comes back into service because something caught a slipping repair before it turned into a week-long delay.

"The win here isn't that the software is smart," Carlson said. "It's that our customers stop thinking about repair coordination at all. It just happens, and they get back to running their fleet."

Where this goes next

ServiceUp's stated goal from day one has been that vehicle repair runs itself. This launch is the biggest move toward that yet, an agent on every step of the process, learning from every repair it touches, needing less from a fleet's team with each one that comes through.

Fleet repair that runs itself. That's the idea. As of today, it's also the product.

To see it in action, schedule a demo at serviceup.com/demo