Cloudera Evolve 2024 | Ryan Garnett, Halifax International Airport Authority

Clips
News
More from Cloudera Evolve 2024

Ryan Garnett

Senior Manager Business Solutions

Halifax International Airport Authority

Three insights you might have missed from the Cloudera’s Evolve24 event

As the digital world grows more complex, organizations find that hybrid data management is the key to unlocking new possibilities. Bringing together diverse data streams and harnessing their full potential is no longer a luxury, but a necessity for staying competitive. The ability to blend proprietary and cloud-based data is helping organizations turn raw information into actionable insights, allowing them to fine-tune their AI capabilities and push the boundaries of what’s possible.Unified platforms such as Cloudera Inc

Enhancing air travel efficiency through data: A Halifax Airport case study

In an era where airport delays seem unavoidable, Halifax International Airport is setting a new standard for operational efficiency by harnessing data for accurate decision-making. How has the airport overhauled operations, using predictive insights, to improve overall customer experiences?Halifax’s Ryan Garnett talks with theCUBE about how airports use data for predictive insights.“I did an assessment, and we looked at it, and the big thing for me is I needed to own my data,” said Ryan Garnett (pictured), senior manager of business solutions at Halifax International Airport Authority. “I didn’t want to be data-locked

play_circle_outline Unveiling the Talent: Speaker Introductions at Cloudera Evolve24 and Ryan Garnett's Data Analytics Career Journey

play_circle_outline Overcoming Data Analytics Challenges at Halifax Airport with Cloudera: A Transformative Transition

play_circle_outline Use of data sets to predict future passenger traffic and streamline operations

Info
Transcript

Ryan Garnett, Halifax International Airport Authority

Ryan Garnett

Senior Manager Business Solutions Halifax International Airport Authority

Ryan Garnett, Senior Manager Business Solutions at Halifax International Airport Authority, shared his experience with Cloudera at the Cloudera Evolve24 event. Coming out of the Covid slump, the airport needed data-driven insights to navigate the recovery phase. Initially, they struggled with unreliable dashboards from a professional service company before switching to Cloudera. With Cloudera, they automated processes, improved predictive analytics, and saved 40% in costs.

By leveraging data from departing passengers, they were able to predict future p... Read more

explore Keep Exploring

What is Ryan Garnett's background and career experience in data analytics and how does it relate to his current role at Halifax Airport? add

What factors led to the decision to use Cloudera for data analytics at the airport? add

What data set was initially used to track passenger movement and predict future trends at airports? add

bolt Powered by CUBE AI

Ryan Garnett, Halifax International Airport Authority

search

>> Good afternoon everyone and welcome back to theCUBE's live coverage of Cloudera Evolve24 here in New York City. I'm your host, Rebecca Knight, sitting alongside my co-host and analyst, Bob Laliberte. Bob, let's talk about air travel.

>> Sounds good to me.

>> Had any delays lately?

>> Oh, yeah, of course. When don't you?

>> Well, exactly.

>> Unless you're at Halifax Airport. No delays there.

>> And you ar using data-driven decision making.

>> Absolutely.

>> So with that, let's invite our next guest on the show. He is Ryan Garnett, Senior Manager Business Solutions, Halifax International Airport Authority. Thank you so much.

>> Thanks for having me. I appreciate it.

>> Yes, absolutely. So thank you for joining us today. Your background is in data analytics. Why don't you tell our viewers a little bit about your career and also what you do at Halifax Airport?

>> For sure. My career has been a little bit different than most people. I've traveled around the world. I started in the geospatial area making flight simulators, so aviation in a different way. So working in Canada for military industrial companies, Bell, a subsidiary of them. And then I traveled around through United States working in geospatial, living in India, China, the Middle East building what we would call digital twins today, but 3D visualizations to do management decisions and help you understand how to run large geographic areas. And then went to grad school, moved to the city of Toronto, ran data there, set up data science for the city of Toronto, and then moved back to Halifax where I'm born and raised. And now I'm at the airport and bringing all those different types of backgrounds around public-private regulations and bringing it in to make a driven aviation experience for our passengers.

>> And you've got a great story to tell. I really enjoyed when we were talking a little bit earlier, I'm wondering if you could share, obviously you're here at the Evolve24 Cloudera event, so you're a Cloudera customer. But I was wondering if you could share a little bit about this was our prior environment and then why we chose Cloudera and this is the environment after.

>> Yeah, I love that story. It's really fun. So I joined the airport just over a year and a half ago. In that timeframe, we were coming out of Covid. And I don't think it's a surprise that the aviation industry had a pretty tough time. So with every downturn or opportunity, that's what data came through the airport. So people need to know when are we coming out of recovery? When are passengers coming back? What's our revenue look like? And that really was the for data analytics at the airport. So having no people there, we outsourced everything. We used professional services, let's build this, it's a project, things of that nature. And they wanted two different types of products. How are we doing for people? And then how are we doing for revenue? So the D were dashboards. So they went with a professional service company to try to pick all these different products and different solutions and things of that nature. And unfortunately, it never really panned out well. They took it as a project and it only got off the POC level. And I don't know about you, but in my career, demos turn into production overnight. And so surprise, surprise, things didn't work the way that they had planned. Things were going down. It was unreliable. Every time you wanted a change or problem, you had to go back to the consultant, it could take six weeks, get out your checkbook, sign new things. So it wasn't a great experience. And during that timeframe, it was really important. The airport was turning off air conditioners to try to stay afloat, right? So it was a big deal and it's going to our board. So when I came in, I was like, "We can't stick with these companies." Not to mention, they weren't really industry-standard style companies that people had ever heard of. So how do I attract talent? How do I build something? I don't even know it myself. So how would I come in? So I did an assessment and we looked at it and the big thing for me is I needed to own my data. I didn't want to be data locked. And I had worked at some other places and we used some hyperscalers in some other big companies, and it was always going to be costly to come out. I'm a proponent of open source. I want to feel like I have a little bit of my own control. And knowing that we're in a regulated environment, I absolutely have to think about audibility and governance and pieces like that. And I came across Cloudera, and Cloudera was something that just worked for us. I also was really a bit concerned if I'm being honest with it, because Cloudera is known for big organizations. And I'm in an international airport, but in Canada, it doesn't mean we're Dubai or anything like that. So we're tier one and my team's never going to really grow to be that ginormous team. We're two people; realistically, one. I'm one and a half. So we weren't sure if that was going to work because you always hear Cloudera with banks and big institutions, 500s, but we are coming up on our one-year anniversary on Saturday. So we've replaced it and it's our entire platform. Sorry for the long story here, we went from two products, two dashboards and two tables that were manually generated, emailed back and forth to a fully automated flow that is telling us when there's problems, over 100 tables running full time. We've never had an outage, we've never had any down... We have, I think, 50 products that our consumers use. We have self-serve people running into it. We now have AI doing predictive models when everything was like, "What happened last month?" All those reactive types of pieces. And the cost savings, we saved 40% alone by going in this direction.

>> So I want to get right into this.

>> I know there's a bit there.

>> Well, this is the thing: we all go to the airport. I mean, most of us do. We need to travel for work, travel for play. And so understanding how to make the travel experience better for us passengers as well as also predictable for the workers, this is such crucial... I mean, everyone can grasp how beneficial this is. So talk a little bit about the kinds of data that Cloudera began to collect for you and what you were seeing, particularly as you were coming out of the pandemic. It was a real game changer experience for not only aviation but for a lot of industries.

>> So at first for us, it was one data set. We just needed to connect to one thing. It was truly started at one. And in Canada, we use a federal organization called CATSA. It'd be the TSA in the United States. So all about departing passengers and every time you've got your scan, your boarding pass, boop, boop, boop. So we'd know how people are moving around. And that was just like, "What happened yesterday? How many people were there? What does Friday look like versus last Friday, this year versus that year?" So we can kind of get an idea... And this has a huge piece around finances, right? Is our credit rating available to go up? Can we make our operational development all these pieces? So that was great. We got it in and we started working like, "Well, we can do a lot more with that data now. Now that we've answered those questions and people are a little off our back, what else can we do and into the passenger experience and the operations?" This one data set, just this one and then I can talk about all the other ones if you're interested, this one data set we use to actually predict what's going to happen in the future. So we know what has happened. We have years of data and we can classify them. We use different machine learning algorithms to do the piece so we can predict what's going to happen tomorrow, what's going to happen in a week, what's going to happen in a month. And that sounds obvious and logical, but when you have decision-making happening in different departments like our commercial department, they're not data scientists. That's not who they are. So they use what's called a load factor. It's like if you have this many seats, we think 80% of the seats are going to be sold for the whole year. So you're working with these very big variances where we're like, "Well, but we can predict a lot closer to that and we can start seeing these spikes that you're not seeing and integrating them and connecting the new data sets like the events that are happening in our province." So we partner with Discover Halifax and they're part of the commercial scene of conferences that are coming in or sporting events. We had the Junos, which is our big Grammys in Canada, that are coming in. We know all the number of people that are coming like, "Well, that's going to influence the model. 10,000 people are coming in this week. That's totally different than what you saw last June or last month." So we're able to bring it in. And then we took it a little bit further. So when we were talking off camera, one of my big metrics is the number 500 million. And it's like what's 500 million? 500 million is what we think we need to spend to build a new terminal in Halifax, which is ginormous for a small airport. For anyone, $500 million is a huge investment. So can we wring what we have in our infrastructure? We've worked with our engineering team, we call them planning and infrastructure. They were like, "When did people show up?" We were like, "Wow, I don't know." We don't have sensors walking around. We don't do that. We were like, "What can we do with, again, that one data set?" So this is the story of one, that data set. We were like, "Well, we know when people scan, right? We have that boarding pass." And we worked with it and now we're giving show up times like to the 10 minutes, this is the number of people that are showing up in 10 minute intervals. And then we know the capacity that CATSA/TSA can scan and get people through. So we can start predicting how long it's going to take and how long a wait line is and how big it's going to fill up our area. And that allowed our planning and infrastructure team to be like, "Oh, we can build one new line this year and do the process rather than three, saving millions of dollars. We can kick the can, so to speak, down the road."

>> Maintain the that people are expecting and still stick to the work too.

>> Reduce that anxiety and frustration of traveling that you can have. And that was great. So we were like, "This is cool. This is super great." And we got validated about two weeks ago from the VP of that group, and they told us about the number one day, engineering works on your absolute busiest day of the year. And we were off by seven people in our predictions. So not only is it great that we did that with one data set, but we are now building the trust within the organization that we can do more than just what happened last month, do a load factor of 80%. So it gets into these other operational factors and we get into all these other data sets that we've brought in. And I'd be more than happy to talk about all those other pieces if you're interested.

>> Well, it was interesting you said our data set, our data set, but it's not one data set. You are bringing in and abstracting lots of different data that you're able to unify and then present up as a single entity that the company can then go in and get this information.

>> Absolutely. And within that Cloudera environment, it's our platform. It is everything. We break stuff out into domains and our domains are around flight information, so schedules. We pull it from six or seven different locations. And then we have our passengers, we have our environment finances, so all of our billing information that we pull in, concessions. So square when you go get a purchase, we know all the transactional... There's so many different data sets that we pull in there.

>> Oh, that's great.

>> That came from the desire of one, but we know we can do more than that. And having this platform, it's really allowed us to do that with a team of one person.

>> The point that you made about how the commercial group is able to use this data, they're not data scientists. They're not trained in how to pull the granular data and then look at it to detect patterns. But in fact, when Cloudera makes it so much easier for you to be able to, first of all, trust the data, know that you're extracting the right insight and then be able to take action and then refine the action if the action wasn't quite giving you what you needed.

>> Absolutely. And commercial's a great example. So we have some people that have moved into those roles and they're there to do some reporting and help out. But to your point, they're not coders. I know I'm not going to teach my HR generalist and our commercial person to be writing in Python or R or SQL. It's just not going to happen. So we work with them to build our products and having a platform that's so dynamic and capable and trustworthy, we build fit for use data. So we go through that evolution and we build these products, these new tables that one of our analysts can pull down and she can pull that down herself whenever she needs it. It's automated, it's trusted, it's there every single time, and we're saving her three days a week from what she was doing manually. So those three days now can go and solve other business problems that they weren't able to tackle. So it's really provided value that I wish I could put a single numeric piece to.

>> Yeah. It's hard to do that when you're able to derive so many different value... There's cost avoidance, there's cost savings, there's the enhanced experiences which might drive additional revenues and support that you're able to get from it. It sounds like you've done a really nice job of nailing down the model that's required for an airport.

>> Absolutely.

>> It now could be replicated or leveraged in another area.

>> Absolutely. I had some of those experience before working at the city of Toronto, and I'm going to get right into that, and that's what I was really conscious about. When I was at the city of Toronto, I was running open data. So it's a data platform to give stuff away. And what we constantly heard was our peers in other municipalities just didn't have the resources. They don't have those capabilities, but they can have those experiences. So it really stuck to me. If we're fortunate enough, we should be trying to build this for other people. And that's what we were taking into Halifax Airport. It's like this is a model, this is a service, this is an approach that can be used anywhere. And in Nova Scotia, in our region, Atlanta, Canada, we're 2,5 million people across four provinces. So one of our local peers, Charlotte Island, beautiful province of Prince Edward Island, they don't have that capability, but they have the same questions. They use the same data sources. So is this something that we can set up and Cloudera itself can support it. You don't need to have this massive team. You don't have to have this massive investment that everyone thinks you have to have in an enterprise data solution.

>> So what is the advice that you would give to other airports or other-

>> Give me a call.

>> Yeah.

>> I think it would be focus. So I do have to give a lot of props to our board and to our executive team is that they're focused. They weren't following the shiny objects.. We want to get there, don't get me wrong, and we're doing some shiny fun stuff. But the focus is build your foundation, have a few questions. These are non-negotiable questions. These are the ones that provide the most business value. And then once you get there, people are going to come to you with questions. You don't have to have all those pieces and you start fighting them off, so to speak. So the big part, if I were to reiterate, is have focus and then build the foundational piece first because that takes care of everything else and it empowers and enables the shiny beautiful things that we all want to play with.

>> And then the foundational layer is the data that you're able to bring up to then create and enable those business decisions to be made.

>> Absolutely. And without it, we're right back to where I was beforehand. Is it down? How long does it take? And then the trust erodes away. And then you're like, "What investment is this? Was this a good investment?" So you really got to focus in on those few things very early on, and then it just takes care of itself.

>> I have to ask though, what are some of those shiny things? When we have you on a Cloudera Evolve25, what are you going to be talking about? What cool shiny new things have you done?

>> Yeah. So obviously everyone wants to talk about gen AI and self-serve and empowering people. So we're doing that. We're testing it in. And self-serve is a little bit different depending on the type of education that somebody has in their background. So we want to put it in there. But the big part that we're really interested in is going back to some of your parts as a passenger, is travel anxiety. It's a real thing for anyone who's traveling, "How long? How's it going to take me? Do I need to be there? What can I bring?" And I hear the stories even from my friends. So we want to empower people to know how long their journey's going to take, what's part of it and make it really, really, really simple so you don't have to be a data person. You can be like, "These are my steps. Here's all that information. And I can take that and make my own informed decisions. Hey, if it's going to take X amount of time, now I know. Now I know and I can work with that. Or I could kind of modify ." And for us, that journey starts when you open your door, when you travel, when you're coming to the airport. Is there a storm that I need to be aware of? Is there traffic? Is there construction? All these things. So you don't have to monitor five or six different platforms. You can just come to one and be like, "I'm going to monitor this." And then if they want, if they choose, they can sign into it and then we can send them notifications to be like, "Hey, so you're aware, we're not recommending, but so you're aware, this is what's going on in your journey."

>> Ryan, I love it. I love it all. Empowering passengers to have a better experience with data. Thank you so much for coming on theCUBE. That was a really fascinating conversation.

>> Appreciate it.

>> I'm Rebecca Knight for Bob Laliberte, stay tuned for more of theCUBE's live coverage of Cloudera Evolve24. You're watching theCUBE, the leader in enterprise tech news and analysis.