Tripian is Disrupting Travel Industry with True Personalization

Jeff Kischuk 

Thank you. Nice to meet you. Hessie. Tripian is a B2B travel service. Where we. There are recommendation service that uses AI and data for services to give really good personalized and destination recommendations for travelers. We’re really helping travel companies fill the gap with personal recommendations and activities in the destination. 

Hessie Jones 

OK, so when I think personal. Because I came from the direct mail world, so everything was personalized. To what extent these days, especially with artificial intelligence. Does travel become personalize? 

Jeff Kischuk 

Well, it’s that’s a really interesting thing when we talk about personalized, I think we’ve all been chasing the definition of personas for years of how do we understand how do we create personas. And one of the challenges is I always use myself as a use case where when I travel, I’ve got 6 definite personas. When I travel, whether it’s business, personal, friends, family, and I behave differently when I’m in a destination. So what we do is we try to identify that persona, help you self identify, help you understand who you are, capture. What you’ve done to when you’re traveling in that persona, be able to make those decisions and do the kind of activities you want so you don’t get recommended to go to a water slide park when you’re traveling on business. And a lot of that persona, the recommendations come down to how we make the recommendations, whether it’s a close referral, whether it’s community based, whether it’s data from your machine learning and what you’ve done. Or whether it’s what I call a, you know, delightful discovery, where we’ll recommend you something that you otherwise wouldn’t do. But it’s because we think you’re in this destination and. And we have a good idea that you might like that type of restaurant, so we recommend it to. 

Hessie Jones 

You OK? So you said a lot of. Stuff there, yeah. So I’m looking at things from a little bit different, different perspective because it’s about making a relevant recommendation to the person who wants a certain experience. But you’re also predicting the likelihood of them actually liking something so how much and maybe this is more of an IP question. So how much of the recommendation comes based on? Predictive the predictive side of the model versus what they’ve already experienced. 

Jeff Kischuk 

That’s a very good question and it’s it’s an area we struggle with, not struggle with, but it’s a challenge because if I break it down to, I’ll take it down to data first of all. So either side of that equation, whether it’s highly predictive or being able to search out a specific activity requires. Very accurate. Data. Being able to have the attribution of those points and of those places in order to even do a match, what we find is over the years and even with AI is is the way we use the algorithms. It’s trying to understand that balance and I wish I had that we do it this way and it’s always prescriptive. But the reality is is we allow users to find their own journey. And find out if they want to be more open to recommendations or if they want to be very pro. Creative and it’s a blend of making sure the technology can support how the user wants to interact because one of the challenges we find with some of the AI tools coming out there is they’re too general, too prescriptive, trying to put everybody in the same bucket where we kind of like that balance of humanity and technology. In order to help you make those decisions and make sure our algorithms can steer you. 

Hessie Jones 

OK, so when yeah, that’s a lot of information.  I’m looking at things more from an algorithmic perspective and maybe before I get deeper into the actual algorithm stuff I want you and I’m going to hand you the mic because my arm is getting really tired, but I want. You to talk a little bit about how it and if you’re using a lot of the generative AI technology today to actually either improve experience scale, the technology, what are you using to? Hi. 

Jeff Kischuk 

So when it comes to the current Gen. AI tools, is again I always refer to them as tools because what we look for is how to use the Gen. AI to enhance that recommendation and user a good example of how we’re using Gen. AI is if somebody likes a place, then you’ve gone to a place and you like and you want to be able to capture your own descriptions. You can use Gen. AI to write a review of that place and the language and style that resonates with you, and then take that Gen. AI will put into our algorithm to do a matching. If you say you have a local pub in Toronto. Really like and you’re going to Istanbul. We can use that Gen. AI description against our POI database to do an accurate match and help you find someplace that’s relevant on your emotions and the way you interact with the location. 

Hessie Jones 

OK, so I’m going to ask you about and maybe this is going to touch on the next question about monetization. So when you you talked to a little bit about discovery. In the beginning. So how do you make serendipity type? I guess. Recommendations to the user and how much of that can be potentially sponsor content and how do you how do you make sure that there isn’t like an over indexing on sponsored type content when it when you’re serving that up to your users. 

Jeff Kischuk 

Thank you. And that’s that’s a really good question because that’s something we have dealt with right from the beginning and the way our algorithms are structured is we make those sponsored, those sponsored POI’s configurable. And the reason it’s configurable because as a B2B, if we’re working with the loyalty company that has a point conversion and participating partners is they have a commercial reason to increase the relevance and the recommendation. And again that is in their ecosystem where their users are expecting recommendations from certain trusted points in their ecosystem. So will allow the client to increase the weighting of those points. To allow them to drive conversion to their business ecosystem for customers who don’t want that sort of weighting, then we use our general weighting where it’s all algorithm based and the algorithm takes into account time of day, whether where you are and just what else you’re doing. So it’s not just about recommending a single point, it’s actually how do you orchestrate multiple points in a smart way. That makes sense. So we’re not sending you to a cafe 5 blocks away and asking you to zigzag back. So the algorithm takes more into account than just what we think you’d like. We also have to balance it with your day. 

Hessie Jones 

OK, so how about ratings and reviews? Because that has been, I guess from my perspective influential in driving a lot of consumer activity. How much of that gets baked into the model and how much of that becomes part of the model after a user’s experience on your plan? 

Jeff Kischuk 

So you must have kind of examined our road map from years. Ratings and reviews is something I think everybody’s trying to crack. Back in 2019, we were working with an AI company that was doing sentiment analysis on reviews to and understand accuracy, which ones were human generated and it it’s it’s. You know that’s an area where as a collaborative partner where we like to be part of an ecosystem is we want to work and find that other startup, that other company who’s cracking ratings and reviews for that authenticity. And allow me to be able to input what kind of ratings reviews I like and what resonates with me. So I think that’s an area that’s still evolving and our technology and our platform is built to take that in. But to answer your question, how we currently deal with it, we do a general weighting of it. We we try to not use it as a driving real critical factor, but what we do is we throw out. Some high and low scores. We also look at the trendings in the scores to understand if a place is trending up or trending down because there was so much inertia in these scoring, it’s really hard to trust. So we we do some things in our algorithm to filter it out to make sure we’re at least having some semblance of a place that is legitimate, authentic. And matches your interests. 

Hessie Jones 

So. I’m looking at a different type of traveler. There are my sister and I are very different people. She likes to go on trips that actually have a scheduled time to wake up, go to this place, take a rest, go to the next place, go to the next place. Everything is time. I like to kind of explore on a day-to-day basis what we’re going to do. Today. So. So from that perspective, when you make recommendations, do you take into consideration things like fatigue, pacing of activities like downtime rather than just maximizing the booking experiences? 

Jeff Kischuk 

Yeah, that that’s correct. So we take into account even the accessibility and the pace, even if you’re traveling with children and how old the children are. So our algorithm takes into account walking pace and. Distance between paces. We also have the option for accessibility to choose which methods of transportation you want to go, whether you want to walk, whether you want to take drive, whether you want public transit, ride shares E scooters, we have a system that calculates the different transfer times between it. Getting back to your first question about that spontaneous. Versus Pre planned is we have a new feature that we’ve released this year that we call instant algo and the instant algorithm is for when you do just want to spontaneously go off and try something new and break your plan or you’ve you’ve gone down an alleyway and you went off your plane cuz that’s how people travel, they’re spontaneous and then will allow you to. Reset and get a whole new set of recommendations based on what you’re changing interests or where you are. 

Hessie Jones 

So from a privacy perspective, you’re curating these experiences based on the collection of data, what specific data? Do you collect and which ones? I guess from your perspective are prioritized that will provide the most optimized model for your users. 

Jeff Kischuk 

Yeah. So with that trip in, the fact that we’ve been working with large B2B companies, we’ve been certified GDPR compliant. We’ve been certified PII with places like Visa North America, we believe, and of course we follow a high degree of privacy and personalization. So when we do not like to track any personal information as a B2B, we like to get a unique identifier. From our customer. And then what we’ll do is we’ll work with them or the client to get their profile and get their interest and get from our client if they have a personal profile on. That user or allow the user to generate some attributes, so we have it tagged to an obscured ID, so we don’t know your e-mail. We don’t know your name. We have you in our system as an ID along with as many tags and preferences which cannot be construed as personal information. So it’s a combination. That’s one of the. Fun challenge. As we have working with enterprise companies, is balancing the degree of third party and 1st party client data they have what they know about their customers and then helping them with getting a mechanism to learn more about their customers? Because what we do is by gathering that information of how they behave and what they’re interested in, we are able to contribute that information back into their BI tools so they. 

Hessie Jones 

Cool. 

Jeff Kischuk 

They can then learn more about their guests and how they travel in destination. 

Hessie Jones 

1. So one last question, it seems like you’re already creating, I’d say, an experience where you know the future of travel isn’t really about the individual. Looking and doing all the research beforehand, what does? How does the experience change from from your perspective like? How do you want it to change? Maybe that’s the better question.

Jeff Kischuk 

That’s actually a very timely and brilliant question is. We’ve always been a technology company. We chase the technology solutions, we embrace technology. But what we keep learning and feedback is there’s a very much human elements. And I mentioned earlier about human ingenuity is we believe and when we travel and when we get feedback from travel. 

It’s we believe it’s going to be a blended experience. It’s taking the human contact of the community, the ability to still get recommendations from your friends and family and being able to put them into a tool and a platform that you can manage and get recommendations and make those decisions or take the decisions that are made for you. So our personal belief is again about using technology to advance. Human interaction, especially when people travel, there’s, you know, people are always going to travel. We believe when people travel together. One of my concerns about some of the AI travel planners is they’re purely A1 to one experience that you’ll get an AI tour guide and the earbuds in your ear, and you’re not interacting with anybody. You’re not with your sister, you know, discovering and learning places. So we believe that, you know, the future of travel is going to be much like the way humanity has progressed. It’s always going to be social. And needs to be interactive, but again, we all chase optimization and we want to help with that. 

Hessie Jones 

Actually I have one more question. You are raising money. Can you tell us a little bit? About that. 

Jeff Kischuk 

We’re at a really interesting point we’ve, you know, been around for five years. So we’re well funded. We’ve had really good, you know, Angel investors and seed through the point. But where we are now, we’re at. A. Scale up and it’s getting to that point where we have. Customers, we have a proven mature product and we’re looking for that strategic partner to help us get to that next step to do more expansion of of enterprise level and make sure that our organization can scale both in port sales and customer service to support those enterprise. Customers, again on a mature platform. So we’re at a really exciting opportunity of a proven company with a mature product ready for that next scale. 

Hessie Jones 

Well, thank you so much and I wish you luck. I’m sure you’re going to disrupt travel the way the way it’s meant to be in a privacy preserving way, right. Thank you. And for everyone out there that we are tech uncensored and we’re going to be a collision all week. So in the meantime it take. Have fun and take care. Tech uncensored and Altitude Accelerator Podcast does not constitute a recommendation for any organization, product or service. It is produced and distributed by Altitude Accelerator for more tech concepts through content. Please subscribe wherever you get your podcasts or visit us at altitude accelerator dot com.