We love the travel space. Our team at MVL has a long history of experience and passion for the travel vertical.
We have been founders of several travel startups going back to the late 90s, including Farecast, which pioneered the use of machine learning to predict changes in airfare and hotel prices (acquired by Microsoft in 2008). Madrona incubated Farecast and has a long history of investing and supporting innovative travel startups, from Spotnana, which is focused on rebuilding the technology infrastructure for the travel industry, to, most recently TrovaTrip, which connects creators and communities directly with tour operators to offer unforgettable group travel experiences.
We’re always looking for ways to push the boundaries of traditional travel, particularly if they align with emerging powerful technologies like machine learning and AI.
The travel space is ripe for innovation because the current experience for both average and frequent travelers still leaves much to be desired. Despite living in an age of customization and personalization, many travelers still find themselves pigeonholed into one-size-fits-all user experiences that are cumbersome and overlook their preferences and specific needs. Any problems arising during the travel experience will likely lead to frustrating phone calls with long wait times (oftentimes in an airport after a canceled flight), taking up valuable time.
Venture-backed startups require a disruptive opportunity and strategy. We’ve been scanning the travel horizon for the last several years and haven’t seen a lot of opportunities for significant disruption - until now.
Generative AI technologies' uncanny natural language understanding abilities will change the game. Large language models (LLMs), AI “agents,” and conversational interfaces are poised to revolutionize how we experience travel today, ushering in a new wave of travel startups focused on personalization, intelligence, convenience, and efficiency.
Let's delve into how these groundbreaking advancements set the stage for a new era in travel.
AI agents and LLMs excel at understanding the ambiguity and idiosyncrasies of human natural language. They can tirelessly sift through countless travel options, reviews, and pricing data to tailor travel recommendations that align with each traveler's preferences and requirements. Whether it's finding a hidden boutique hotel, a flight with the most comfortable seats, or a vacation package that caters to unique interests, these AI systems can curate options with a level of personalization previously unattainable. This customization extends to different aspects of travel, including destination choices, dining preferences, activities, and even suggesting the best times to visit attractions based on historical crowd data. They can continue checking your favorite hotel to see if a room becomes available or obtain credit if a flight price drops after you purchase it.
The technology will learn, adapt, and remember from interactions, enabling more personalized responses. For instance, it can remember a user's preferred airlines, meal preferences, or seat choices and automatically incorporate them into future search results and suggestions. They can also analyze customer interactions to understand common issues and preferences better, enabling them to provide more accurate and helpful responses over time. These personalized experiences will make each interaction more relevant, efficient, and enjoyable.
From a technology perspective, travel can be deceiving because the top-level information a traveler needs to convey is very structured and standard (origin, destination, departure, and return dates, etc.), fitting nicely into a webpage form. And yet, the devil can be in the details when it comes to articulating preferences that make the trip pleasant and smooth. This is why, more often than not, a traveler needs to interact with the human to understand their intent fully.
New natural language interfaces will significantly shift how travelers interact with technology, simplifying searching for flights, booking hotels, planning itineraries, and articulating and remembering traveler preferences. Natural language communication will allow travelers to interact with devices and services more intuitively and easily.
Instead of navigating through lengthy forms, complex menus, or repetitive search queries, travelers can now simply share their requests in natural language. Travelers will be able to type or say, “I need to travel to Atlanta for a meeting next week,” "Find me a flight from New York to Paris on July 15th," or "Book a hotel near the Eiffel Tower for next weekend," or even, “I’d prefer Delta airlines, but if Alaska is available that would be fine as well.” Notably, the inherent ambiguity of natural language will no longer be the barrier it has long been when trying to communicate with technology.
Dealing with last-minute changes and real-time updates are other areas where state-of-the-art natural language understanding will shine in the travel sector. Travelers can receive up-to-the-minute information about their trips just by asking. For example, imagine being able to simply ask, "What's the status of my flight to Tokyo?" “When should I leave for the airport?” or “How long will it take me to get through security?”
In customer service, conversational interfaces coupled with AI agents and LLMs will revolutionize the experience by providing round-the-clock assistance. These systems will be capable of handling a wide range of queries – from basic questions about baggage allowances to more complex itinerary changes or last-minute travel disruptions. It can be incredibly stressful when you arrive at the airport to find your flight was just canceled. Instead of simultaneously standing in a customer service line, searching for other flights on your phone, and waiting on hold to try to talk to a human operator, a traveler will be able to simply communicate, “My flight was just canceled, and I need help finding another flight home tonight!” to an AI agent, which can then immediately get to work helping them get home.
These systems will also be instrumental in quickly and simultaneously handling large volumes of customer inquiries, particularly valuable during peak travel seasons or unexpected disruptions. By efficiently managing these high volumes, they will reduce the strain on human customer service representatives, allowing them to focus on the more complex and nuanced customer needs. The speed and accuracy with which these AI systems operate will also significantly reduce wait times and improve overall customer satisfaction.
Below is a fragment of a virtual travel agency architecture. Each software agent, powered by an LLM, is instructed (prompted) to perform specific and specialized tasks (e.g., flight search, corporate policy rule enforcement, booking, etc.). An orchestrator agent coordinates work between the agents, accomplishing various tasks on behalf of the traveler. The flexibility and modularity of this new type of architecture can dramatically lower the system's complexity and reduce the brittle rules and complex code that current systems need to maintain. It also promotes the scalability and flexibility of the system. Travel domain experts can make changes or additions to the system without requiring software engineers. This alone can make travel information systems more agile and adaptable to change.
In the dynamic landscape of travel, the integration of large language models, AI agents, and predictive analytics isn't a mere enhancement; it's a seismic shift that redefines every facet of the travel experience, ushering in an era where personalization, efficiency, flexibility, and innovation converge to create journeys that transcend expectations.
If you're a visionary founder thinking about the next generation of travel products, we’d love to chat with you! Learn more about MVL by delving into our manifesto here.
We are with our founders from day one, for the long run.