Carlos Romano, Director of Product Management, leads the design of Garage. Here, he talks through how AI fits into fleet management and what AI can make possible for the people who run fleets.
How does AI play a role in what we do at Inspiration? What do people get wrong about how AI supports fleet management?
AI is part of how we work every day, not a side project. The internal tools we’ve built are genuinely cutting edge, and they change what our team can do. Each person at Inspiration does the work of several. Our team learns from every customer engagement, and that learning makes each person faster the next time.
Our customers feel it most. When we work on a project with a customer, what we learn stays with us: how their fleet runs, the processes they follow, their preferences, and what we agreed on and why. It doesn’t depend on any one person. Whoever on our team picks up their next call or request starts with the full history. The customer never has to explain things again or watch something fall through the cracks during a handoff. They’re working with all of Inspiration, not just one person.
Otto™, our AI assistant in Garage™ 2.0, brings that same thinking into our fleet management platform. A fleet manager asks a question in plain English, and Otto answers from their live fleet data and can take the action for them.
The usual mistake is thinking AI in fleet means a chatbot on top of a dashboard, or a set of rules with a new label. Take utilization: a hard and fast rule for how to determine utilization across every single fleet is wrong. One fleet runs at 45 percent normally and another runs at 85 percent. A driver who usually does 300 miles a day and logs 310 is fine. A driver who usually does 80 and logs 310 needs a look. Every fleet is different, so one definition can’t fit them all. Our tools combine AI with real flexibility, and we set them up around your fleet. We don’t expect customers to change their processes and tools to match our preconceived notions of how a fleet should work.
What’s the most important insight you’ve gained about AI through building Garage?
The model isn’t the hard part. Context is. AI is only as good as what it can see, and in fleet that context is scattered: telematics in one place, fuel cards in another, maintenance with a third vendor, and leasing with a fourth.
Garage is built around bring your own vendor, or BYOV. Customers keep the vendors they already work with, and Garage connects them. Every vendor we connect adds to a complete picture of the customer’s fleet across all of their vendors. That’s what lets Otto give an answer you can trust instead of a guess, and because Otto runs with the logged-in user’s own permissions, it can never see or do more than that person could.
Which AI capability in Garage are you most excited about, and what does it let a fleet manager do that they couldn’t before?
I’m excited about two things.
The first is that Otto takes action instead of handing you a report. Our shorthand for it is “every action is a sentence.” A fleet manager types: “Find unassigned vehicles garaged around Dallas that are up to date on maintenance.” Otto searches by location, assignment status, and service history in one go. Then they say, “Assign that van to an available driver in Dallas and enroll them in home charging reimbursement,” and it’s done. Before this, that was an export, three screens, a spreadsheet, and probably an email to our ops team. Now it’s a short conversation.
The second is that fleet managers can build their own views of the parts of the fleet that matter to them. Every fleet manager’s job is a little different, and a dashboard we design will never fit all of them. With Otto, a fleet manager can say, “Every morning, show me the top three things I need to do today,” and decide what that means for them. Maybe it’s leases about to expire, overdue maintenance, or drivers who need a vehicle. Otto helps them define it, builds the view, and keeps it up to date. They’ve never been able to do that until now. They used to get the screens we gave them, and now they can describe the one they want.
Together, those two things move Garage from a system that records what happened to one that helps you decide what to do next and then does it with you.