Innovation Roadshow: Flyr Labs

Transcript

Matt Brown:All right, well, thank you so much. My name is Matt Brown. I'm the VP of Growth for Flyr Labs. We're a data science organization that has been in the airline space now for over 3 years, based in the US, San Francisco, but offices globally. And I'm going to talk to you today a little bit about how we've taken a fresh perspective and approach to helping airlines and other travel partners solve some of the commercial issues and commercial problems that they're facing in everyday challenges that this industry presents. So the problem statement that we believe persists, and you've heard it consistently across a lot of speakers today, is that the system that exists today, the tools, the processes, the methodologies, really haven't been able to keep pace with the industry and our ability to adapt to it to optimize commercial decision-making. And so Flyr really started, frankly, as a few people in a garage in San Francisco who had been spending quite a bit of time focusing on how other industries were solving problems using artificial intelligence. Now, we hear that word everywhere, right? AI is a buzzword. Everybody's using AI. It's solving a lot of problems across a lot of sectors. And I want to talk to you today a little bit about how the application of AI can dramatically improve and is improving for our customers' commercial outcomes. So first, let's look real quick at— oh, I'm in a bad spot. Let's look real quick at a study here from McKinsey. And what's very interesting about this is they effectively have said that AI on its own represents about a half a trillion dollar opportunity. The travel industry alone. So there's this tremendous amount of trapped value or trapped revenue that's sitting in the travel space that could be unlocked by AI. And what's even more interesting is if you double-click on that, the area in which AI can have the most impact to drive that value sits in marketing and sales. About 90% of of our ability to generate offers, orders, continuous pricing for the traveling public sits with the ability to be improved with AI. And so, you know, the question is why? What's so special about AI that allows for these dramatic improvements in commercial decision-making? And I like to think of it like a flywheel. So the beauty of AI is, unlike us humans, an AI is constantly learning. It's constantly training. So at the point that it has more data or more ability to inform better decisions, it learns from that. So let me give you an example. I've got the new iPhone myself, and if you think about what Apple does, every year to release a new product. Imagine the amount of work, the design, the need to develop a new chip, new capability, the requirement to manufacture this device every single year is a monumental effort. Now imagine if you could automate that, right? Imagine if you're a service industry like an airline that has access to just an enormous amount of data. And that data can be continuously learned from in real time to drive better outcomes for airlines or for other travel spaces. That's the beauty of AI, is it's continuously learning and continuously providing better results. So real quick, some grounding in AI. Artificial intelligence has been around for a very long time. In fact, it dates back to the '50s. So it really means anything or any machine or computer that can sense or that can reason or that can— can adapt to a specific output. Machine learning is an advanced form of AI. And this is one where algorithms tend to improve their performance over time. This is the continuous learning component. And deep learning, which is a form of machine learning, uses neural network technology. Think of it a little bit like sort of reconstructing the human brain in order to drive better decisions. Or present better options. And that's exactly where we've been investing as Flyr for many years now. And you can kind of think of it a little bit like navigating your car or driving a vehicle in a city that you know well. You know, Google Maps, Waze, Apple Maps, they do a great job at this because what they do is they create context from a whole bunch of real-time data. So they say, hey, you want to get from point A to point B. Now, you can— you normally know that during rush hour you shouldn't take Main Street or you shouldn't take a certain large highway. But it calculates all of that context and presents an optimal route based on all the considerations that it has to have. That's exactly what AI can do when you have access to tremendous amounts of data. That can real-time provide an optimum solution for any user if it's appropriately set up. And in fact, we think that we're in pretty good company because we're using deep learning technology much like folks like Google, Netflix, Tesla, some of the other folks that you see on this screen. And it's that community effect, it's that ability to learn about how to do— how to apply the technology better that we are able to take advantage of when we're driving results for our airline customers. So for us, the ability to take in tremendous amount of data and make sense of that data sits— underpins the ability to really change and improve commercial outcomes for our airline customers. So this is just a small sampling of some of the data that we're using inside of our platform to be able to drive results for our customers. But I want to sort of talk a bit about how all of that comes together. So we build a platform that takes all the data that I just showed you, plus new data as it comes in and as it arrives, and in real time outputs forecasting. So if you can forecast better using AI, you can traditionally make much better pricing decisions for revenue management, and you can optimize total revenue for your aircraft or for your operations. And it does this in real time, and it does this with much superior results than traditional capabilities, which tend to look sort of in the rearview mirror. They look at the past to predict the future. And we all saw what happened during COVID when that type of methodology was relied on to be able to determine the best next steps to move forward. So the platform for us consists of a whole set of data, An ability to drive better forecasting leveraging that data, ability to drive continuous real-time pricing outputs, and then also provide intelligence or business intelligence to make better and more informed decisions. And really for our end goal, it's about maximizing revenue for our airline customers. So it's taking into consideration not just fare revenue management, but over time ancillary revenue management, cargo revenue management, etc., to provide an end-to-end total revenue optimization using the same platform of artificial intelligence. And don't just take my word for it. We're— today we're managing over $15 billion of our customers' revenue, and that's with multiple airlines of all shapes and sizes across the globe. And here's the results that we're seeing. And these results are impressive because simply by leveraging your data and unlocking the value of that data, we're seeing 5 to 7% revenue uplift, which equates to usually 4 points of load factor increase, a 5 to 10 times reduction in forecast error, and about 90% reduction in the amount of activity a human analyst has to apply when they're managing the markets for pricing considerations on a daily basis. And what's interesting is this is all of our data comes from comparing capability against legacy systems. So if you take a look on the left here, you see the legacy system that this specific example is compared to versus our system, and you can see that it's a pretty significant revenue increase For that specific airline. If you look at forecasting, now this chart's a little bit tricky, but on the x-axis, as you move your way to the right, that's the number of days out from a flight departing, and the y-axis shows sort of the mean absolute percentage error. And across the top is the legacy system in place that we compared our forecasting results to, and along the bottom is the forecasting capability of the Flyr system. And so what's really impressive is not how accurate the forecast is closer to the departure date, but it's really how accurate the forecast is further out. And that's really, really important in commercial decision-making because it opens up a whole bunch of different capability for us. And that's really how our company has evolved. So we started working with revenue management under a common platform. We've moved to including ancillary revenue management. So if you can now think about the trade-off between a seat price and an ancillary revenue opportunity and start to get a total revenue picture, it's a really compelling single pane of glass usage of the platform. We've moved into personalized offers, schedule and planning, marketing capabilities. If you can know 120 days out which markets need demand stimulation and apply that, It's very, very easy to do so using the system. And then more recently, air cargo. Obviously a very important area these days, but being able to forecast the ability to better revenue manage your air cargo is a massive need among our airline customers. I'm going to end on sustainability because I think that's the theme of the conference and it's the theme of almost every aviation conference we've been at. And I would just say that as a technology provider, we often get sort of the get-out-of-jail-free card to talk about sustainability. But much like most of the speakers have said today, everyone plays a role in sustainability. So our belief is that if AI or deep learning technology can do things like allow you to make better commercial decisions, it can certainly do the same for sustainability decisions. So if you have a very accurate set of data— sorry, a very accurate platform using data, you can make better decisions about where to allocate your assets. You can put sustainability goals inside the framework upon which you are making decisions. And ultimately, you can do that without having a significant impact on the customer experience. Please stop by and see us at our booth. Thank you so much.

Copyright policy: All transcripts on this site are the copyright of CAPA - Centre for Aviation. Our reproduction policy is as follows: you may quote up to 400 words of any transcript on the condition that you attribute the transcript to CAPA - Centre for Aviation and link to the original video page. All other use is prohibited. While we aim for 100% accuracy in the transcript, there may be some minor transcribing errors.