Macy has been working for CNET for coming on 2 years. Prior to CNET, Macy received a North Carolina College Media Association award in sports writing.
Over the weekend, I spent hours searching for flights for a summer girls' trip and came up empty. Every option was either too expensive, landed at the wrong time or had two stops on the way -- which I'm absolutely not doing. I checked multiple airlines, pieced together routes and even considered separate tickets. Nothing worked.
That kind of frustration is exactly what Mindtrip is betting on.
The AI-powered travel platform is launching a new flights feature designed for the kinds of messy, real-world searches that traditional booking tools struggle to handle. Instead of optimizing for simple routes, Mindtrip is focused on the complicated scenarios travelers actually face, where flexibility, preferences and trade-offs all collide.
Read also: Google's New Travel Features Are Here in Time for Summer
Mindtrip already combines conversational trip planning with a visual interface that pulls in maps, reviews and itineraries. With flights, it is extending that system into one of the most time-consuming parts of travel planning.
In a virtual demo with CEO Andy Moss and product VP Abby West, the company positioned its approach as less about speed and more about reasoning. The goal is not just to return results quickly, but to think through constraints the way a real traveler would.
That shift is showing up in how people actually search, too. According to West, many people do not start with a fixed destination. Instead, they describe a set of conditions. For instance, they might want somewhere warm within a four-hour nonstop flight, or they'll ask when they can get to Paris within a certain budget.
Those kinds of queries are difficult to execute manually. They require checking multiple destinations, comparing dates and factoring in seasonality.
Mindtrip's system treats them as a single problem. It samples across routes and timeframes, weighs constraints and returns a short list of options that fit.
"We've very much always focused on the full connected trip -- how you plan everything you need around a vacation, from flights to hotels, to things to do, restaurants, anything," Moss said.
"The use case that Mindtrip flights is really focused on is the more complicated travel cases."
In one demo, West searched for a trip from Washington, DC, to Los Angeles, with a long list of conditions. The trip needed to be four nights in June, return by a specific date, depart before 9 a.m., exclude a nearby airport and include a carry-on. Instead of forcing those filters into a rigid form, the system broke the request into parts, evaluated multiple airport combinations and surfaced a set of tailored itineraries.
Each result came with a short explanation of why it matched the request. From there, West could move directly into checkout to book her tickets.
The level of personalization depends on what Moss describes as "practical data," not invasive tracking. The system can account for things like preferred airlines or whether someone prioritizes nonstop routes. It can also adapt to context, such as traveling with family versus traveling solo and then adjust recommendations accordingly.
"I do think you're going to have a personal assistant [in the future]. I do think you're going to have expert assistants that are really good at flights or hotels and those two things will work together and you're just going to basically have a sort of situation where it's almost like Jarvis from Iron Man combined with Her [to create an AI assistant] that knows you really well and understands you," Moss said.
Flights also required a deeper level of infrastructure than other parts of the platform. Mindtrip partnered with Sabre to access global pricing and availability, and with PayPal to power checkout and buy-now-pay-later options. At launch, PayPal is offering a roughly $50 credit on qualifying bookings over $250, a small but notable incentive in a currently expensive travel market.
Mindtrip is not trying to replace tools built for quick, straightforward searches. Moss is clear that if someone wants a simple one-way flight, existing platforms like Google Flights already do that well. The focus here is on more complicated cases, where planning becomes time-intensive and fragmented.
That focus reflects a broader shift in how AI is being used. Instead of instant answers, companies are leaning into systems that take longer but handle more complexity. Moss believes that travelers are willing to wait for better outputs if it saves them significant time in return.
The same approach is expected to expand beyond flights. Mindtrip is already applying similar agent-driven logic to hotels and is working toward a more connected experience across booking, itineraries and in-trip planning. Over time, that could include more automated checkout flows as people grow more comfortable with letting AI handle multi-step transactions.
Even as airfares rise and the travel landscape shifts, demand has held steady. Moss sees that as a sign that planning tools will only become more important. "I don't think there's ever a time when people have needed to travel more," he said.
The challenge is not convincing people to travel, but helping them navigate an increasingly complicated, pricey system. After my own failed flight search, that pitch feels all too familiar. The problem isn't a lack of options; it's the effort required to sort through them.
