Ep 11: Shikhar Agarwal- Will Cloud & AI Revolutionize Travel?

00:00 - Intro the Shikhar Agarwal and Spotnana

Matt Brickey

Welcome to the Cloud Currents podcast, a place where you can hear insights from experts who are revolutionizing how businesses thrive in the cloud. I am Matt Brickey, senior vice president of professional services at TierPoint, and I'm responsible for our consulting and public cloud business units. This episode is going to explore how a startup navigates innovation and transformation in the corporate travel industry. We'll also discuss how the cloud and other new tech continues to enable their growth. Today, I'm joined by Shikhar Agrawal. Shikhar is a cofounder and CTO of Spotnana, a startup disrupting the trillion-dollar travel industry. He has over ten years of experience in cloud infrastructure, databases, machine learning, and leading engineering teams. And he previously worked at companies like Google Brain and TensorFlow. Thanks for joining us today, Shikhar.

Shikhar Agarwal
Great to be here with you, Matt.

Matt Brickey

Let's start with a few questions about your personal journey, because the industry that you're in is pretty fascinating now, and you're doing some very disruptive things to what would be considered a very legacy kind of type industry. But I want to know what first drew you into technology and engineering as a career. Did you ever have a specific aha moment or a specific person maybe that led you into that.

Shikhar Agarwal

Absolutely. So, you know, it's a funny thing that India, I grew up India, and they say that India you only have options to either become a doctor or an engineer. So that's what they say. But for me, coincidentally, my mom is a doctor, and my dad is an engineer, right? So, I bought them growing up. And at some point, of time, while growing up, my father bought me a computer, like a PC computer, like back in the days. And I was very fascinated by that computer. And it was all windows at that point of time. And Bill Gates was one of my ideas growing up. Right? So, I think that entire the engineering mindset that I was grown into in my family, plus inspiration from Bill Gates, really motivated me to choose computer science and engineering as my chosen field.

Matt Brickey

That's cool. And then you worked in some large companies and some small ones. But I noticed kind of going through your background is that you didn't spend much time in the large organizations before you got into founding companies of your own. Give the audience a little bit of perspective of how you went from, say, being a staff engineer at VMware to only a short years later you were founding your own companies.

Shikhar Agarwal

Absolutely. Yeah. So, my career started with the VMware, and there I realized that's not my cup of tea. It was slow moving, less impact lot of pieces of approvals you want just to make a simple code. And at that point of time, I got lucky to join a startup called Thought spot. We were still like 1015 people was in stealth mode and a colleague of mine at VMware actually referred me there because one of his classmates from university was one of the co-founders, right? And that was like contagious that moss in a startup, right? Things move so fast. We're in a small room, marketing, sales, HR, recruitment, everything is happening in that room. So, you not only learn technology and expertise in technology, but you just learn everything that happened inside the black box called startup, right?

So that was really like, it was a mind-blowing experience for me. And then I joined Google to see, hey, maybe Google is different from VMware in the sense a large company. The culture of Google was very different from VMware. But at the end, again, that impact the pace. That was not something that I really liked. It's great companies, but not for me at that point of my career. And so that's why, you know, I've worked in a startup now let me use my energy in creating something on my own.

Matt Brickey

You could always go back to the bigger companies when you get old and you're a little slower and you have less energy, I guess, right?

Shikhar Agarwal

Absolutely. And yeah, that's what my thought was, that instead of asking questions why to do something, sometimes we should ask question why not? And that's what I asked. Like why not leave Google at this point of time? And sometimes the why not really helps in your career or just in life.

4:45 - The Founding of Spotnana and Industry Disruption

Matt Brickey

What was the most unexpected learning experience you had when you founded Spotnana or any of your other organizations? Just kind of curious. The thing that caught you by surprise that launched you even more forward than you are today.

Shikhar Agarwal

Absolutely. So, there were lot of aha. Moment in Spotnana, like in every part of my career. But if I just focus on Spotnana, one of the biggest learning I had was that as a tech entrepreneur or someone coming from software, you always focus on some hard technical problems, right? That's what you like solving, that's what give you the kick. But what I realized is two things. One, business has to drive technology. Technology is an enablement, and it has to be business. 1st, 2nd, when you are in a startup, you're actually in a people's business, which is very counterintuitive. If someone comes from pure technology background that if I look at my time spent, I think most of the time people might think I'm doing some sexy work of solving hard technical challenges and whiteboarding. And that is part of the role. Right.

But it's a lot more is just people's business. It's not just management, just managing expectations, managing communication, bringing everyone on the same page about the company's goal, short term goals, long term goals. And your same message can be perceived by different people in a very different way. Someone coming from India might perceive it very different from us, or someone from a tech background might have a very different conclusion out of the same message from sales background or marketing background or HR background. Right. So how do we make sure that everyone comes on the same page, and you move the company forward? That's a people skill that for me it was pretty challenging to just realize that it's not always logical numbers or data, you have to take all that into account.

But at the end of the day, you're dealing with humans, you're dealing with peoples, and it is a team which makes you successful, not an individual. Right. And that transition from pure software to now we are a team of 350 people and I'm still learning every day the people's challenges.

Matt Brickey

So, you're chief evangelist too? Not just chief technology officer it sounds like.

So, I'm curious because corporate travel, right? And organizations that assist with corporate travel, some of them are decades old, right? And some of them are brand new. I'm curious, what convinced you that disrupting the corporate travel industry was a worthwhile endeavor?

Shikhar Agarwal

Absolutely. So, first of all, this idea was not mine. This was my co founder's idea. And he comes from 20 years of corporate travel experience, right? So, when I met him at a laundromat in Palo Alto, he told me about this world of corporate travel and I've not traveled a lot in corporate, like as a corporate traveler. I went for some recruitment events while at Google and I just booked on my own and got reimbursement, right? So, I started digging deeper. I went inside Google and saw Google, which is one of the best or most advanced technological companies in the world, they're still using legacy tools for corporate travel. They had all the best tools for everything else, whether it's HR or AI or some cloud.

But for corporate travel, it was still legacy tool like conquer travel and some travel agency behind the scenes. And most of it was done manually. Then I checked with my father. My father has done a lot of corporate traveling as part of GE, Alstom, Hitachi. And he just went mad. He was like, when I booked this travel, this agency did this to my hotel and this to my air ticket and all. And that's when I realized that this market seems to be huge. It's right for disruption. The best companies in the world are still using archaic tools and so I still didn't know about anything about the industry. But the problem was obvious that it's large enough and it needs to be solved by someone. And that's when I decided to leave.

Google started and till date every day I'm still learning new things about travel, which is really exciting.

9:05 - Leveraging Cloud Infrastructure and Technical Challenges

Matt Brickey

So, tie your background together infrastructure and database and machine learning into spontaneous platform. How did that go from an idea to laundromat to you being able to apply your previous experience?

Shikhar Agarwal

That's a very interesting question. So, I think the biggest asset that Spotnana has is that I didn't come from the travel background. So, I was able to bring fresh eyes and fresh experience to this travel world. So, if you look in the corporate travel sector, most of the companies, or the big companies like we call the ABCs of travel, these are like Amex, GBT, BCG, Travel, CWT, like the big companies and all of them are doing, at least if you look at Wikipedia, more than 20 $30 billions of booties every year. That's huge. All of them are demons and proven ways and stuff like that. And there is no one of record data in distance you are working on in distance. So, there is just a mess. And for the listeners you might be thinking that we created new.

Yes, we created something new. But it must be obvious in the sense that this industry won't just like we are healthcare that talking about microservices, cloud native architecture, API native architecture is considered new in the corporate travel industry, consumer travel, you have Expedia and Airbnb, all good tech companies. But corporate travel is very different. And so, when we started Spotnana, what we realized is first we need to bring the cloud native advantages to this world of corporate travel so that we have agility, we have microservices, right? The DevOps is good, and we can have separate teams like all the good advantages of microservice based architecture and cloud native architecture. So one was that other things was how do you make sure that you have a single system of record that powers all the data?

So, whether someone is booking travel which ties to their analytics in real time, like a corporate trying to investigate, how much did I spend? How much of it was in policy or out of policy? How much is my sustainability, CO2 emissions? If my travelers are in Dubai and some hotel in Dubai catches fire, I should know who of my employees are staying in that hotel or in that region so that we can help them. Like Isa, like, you know, some war happens and all those things. So, from security perspective, they need to track their employees. So, all of this coming in one database or in one central piece, which powers entire ecosystem, that was never done before in the world of corporate route, right?

And once you have all this structured data globally at one place, now you can do huge amount of AI ML on top of it, right? One is for agents. Like when you are trying to book a flight and you're not able to do it, you can call an agent, chat with an agent. All that is driven by chat GPT and by chatbots in this model time, right? Similarly, personalization, ranking, analytics and insights of how to save cost on your corporate travel program. So, all that experience came together here in Spotnana to build the industry first travel platform, which is all open API native and so on, AI first secure, and all those things.

Matt Brickey

When you first started, did you have the vision that you were onto something big and this needed to scale globally, or was it kind of an iterative event? Give us an idea of those aha moments where you're like, it's terrifying and exciting all at the same time, where you're like, oh my, they really sold this to this company? This company, this company. I hope it works.

13:03 - Scaling the Business and Navigating COVID-19

Shikhar Agarwal

No, absolutely. So, when I started with my co-founder Spotnana, the opportunity was very clear that the market size pretty big, right? So were very clear that we can potentially create. So were not afraid of the market that were like, there is no market risk, there's execution risk. And execution risk is good because it's under our control, right? Like, we need to work hard, hire the best talent, have good VC funding, like the entire ecosystem, so that we can execute, right? So, all was good. But after three months of starting Spotnana COVID hit, and suddenly, when we thought there's no market risk, market went to zero for travel, right? So, these are the ups and downs that are happening. So, we started Q four of 2019 2020 COVID hits, travel stopped. So were like, hey, were thinking there's no market risk.

And now we have realized it's only market risk. Execution we can still do, right? But the AHA moment was that lot of companies and travel agencies in the ecosystem, because they were not tech first. They had huge employment cost, huge licensing costs for other software. Some of them went bankrupt, they closed someone, their aggregation happened, some agency bought some other agency. Like, a lot of those things happened in the industry, and were still in the build phase, right. We were not in the sales phase, so were still in the build phase so we could build peacefully while in the ecosystem. A lot of these things were happening. The second thing was our enterprises, the corporate buyers, typically if there's a lot of people traveling for them, it's very hard to imagine moving to a new platform, right.

It's a massive change management for them. But in COVID, no one was traveling, and they had a point where they can take a step back and see, hey, what else exists in the system. Are there any startup that we can take a potential bet on? Because no one established. So, it gave everyone that opportunity to think what's out there and to see if they can take some calculated risk on a startup, right. So, we saw that opportunity. Our VC saw that opportunity. So, in COVID we did seed, we just did before COVID but we did series A, series B. Like series B, we did around $70 million. Series A was like $30 million from some of the best investors out there because they saw this opportunity. So, all this is happening.

No travel, it's COVID huge market risk, people still believing so they're funding because they know that this could potentially be huge from demand perspective, other travel agencies going bankrupt, or they're being bought up. The buyers having an opportunity to explore what's out there. So, all of these things are coming together while were just hiring. And we grew from zero to 200 people in COVID or remote hiring because we could not physically meet anyone. So, all this were happening during that time. Right. And then were pretty fortunate that some really great corporate buyers took a bet on us. So that was pretty big. And I'm still not at the liberty to tell, but some of the top ten, like Fortune ten companies are our customers. And in the next few months we'll be telling all these names out there that who all are customers.

One of the Qantas airlines is also a customer. It's not Fortune ten, but this is publicly available. But this is what was happening. Everything altogether.

Matt Brickey

If that happened, right. In COVID, how did you encourage your team to stay with the cause? Because that's about as disruptive as it gets is that you're in the corporate travel industry and all travel stops.

Shikhar Agarwal

Yeah. So that's a very good question. So, what we did was that initial few engineer that we hired like 20 ish, we hired from network, right? Like my juniors from my university or my colleagues. And also, that was still there. The other thing was, and you're absolutely right, like we are a startup in travel domain, in COVID. So convincing people was difficult, but the people that we hired, they referred their friends. So that was again a very big plus. Plus, it just took us that either me or my co-founder Sarosh, conversation with every engineer people that to convince them, hey, this is how much money we have got. This is the team we have; this is the problem we are trying to solve. And so that conversation really helped in selling to these employees.

18:07 - Maintaining Startup Culture and Team Motivation

Matt Brickey

Okay, great, let's switch gears a little bit and let's talk a little bit of technology, right? Is that I'm curious what some of the biggest challenges you faced in leveraging cloud infrastructure or just technology in overall, right. As you're building out a new platform, what are some of the challenges that you guys have faced? And.

Shikhar Agarwal

You know, from outside of it looks like it's all cool and sexy and very easy. It's all known, just leveraging it. First challenge I faced was that the company that I have worked, VMware, Google, ThoughtSpot, they build all the technology in house. So now thinking in terms of using a public cloud and not building everything but using open source software available outside. So that was a mindset change that I needed to have. And what really helped me was just hiring few people in the initial team who came from that background using cloud infrastructure and all. So that was number one. And then definitely then it's a journey. Suddenly you've realized, oh, there are five types of same relational databases available and three types of caches available. So, all those time took something like what are the pros and cons?

One other thing was how married we want to be to one cloud, right? Should we be very native to AWS, or should we be cloud agnostic? And that decision for us was also critical, because now your architecture would depend upon how you build the wrappers when you're connecting to a particular router. So, we are an AWS shop, but we are cloud agnostic because we know at certain point of time some customers might say, hey, we definitely want azure, or in some countries we might have limitation, right? So how do you make sure we are cloud agnostic? So that architecture needs to be, we need to make sure that it's architect from day one. So, this was one part of it. Second part of it was because we are in travel.

We have your passport information, date of birth, like credit card, all the PI information, because you're booking your flights and all on our platform. Security was huge for us, right? And once you go for now, you want sock two, you want ISO, and you want PCI, like all those things. It's not that we are an expert in security and again AWS publishes whole bunch of documents on cloud security and all those things. But to parse it's very hard, right? And what infosec documents you are making. So, security was something else. So again, we had to hire some industrial expert who has been in the cloud. That really helped us. The third thing that happened as we start moving on is the expenses. Like our AWS bill is almost now reaching a million dollars.

And we are seeing, hey, what we're realizing that we have so many environments staging pre-production, QA environment, and every time we spin a new environment the cost just almost linearly increases, right? So, we are not just playing with production but all five, six, seven environments that are running in parallel, which is mimicking the prod environment, right? So, the cost is becoming pretty high now. So now we have to figure out how to optimize on the cost. So, these are some of the challenges, but definitely even I would say the CI CD. What I would hope is that if AWS would have a pretty mature CI CD system like they had their own pipelines, but all of those were lot of challenges even to just do a Ci CD. Like we tried harness and circle CI.

As of now we are using harness, but it's also not perfect, right. Perfect solution doesn't exist in general, but just the native thing provided by the cloud for your CI CD is not mature enough. So, these are some of the challenges that come top of my mind. But I'm sure if we start digging deeper, we will definitely have lot.

Matt Brickey
I'm curious how you're approaching. Sorry. Finish your thought.

Shikhar Agarwal

I'm saying the latest now we are realizing is even Datadog is becoming very expensive and all those things, that's what we are realizing. It's not cheap, it's really becoming very.

Matt Brickey

And the more things that you bolt on, right. They become part of your platform that then you have to either write it yourself or look to other platforms. But that kind of leads me to my next question. I was curious because you mentioned you wanted to be cloud agnostic and you wanted the ability to port your platform to another cloud. And I think that's very wise because you are going to get asked that eventually. I don't want to have relationships with AWS. I'm Microsoft shop, vice versa, whatever. How do you approach that? With reliability, compliance, and you mentioned security. Give me an idea of what your wrapper looks like around that. To be able to move that from cloud to cloud?

Shikhar Agarwal

Absolutely. Principally what we have done is that first of all, for example, if we take something from AWS, we find that it's the same offering available in other clouds. So, for example, if you're using a queue management system, so you have Kafka, but AWS has their native SQs and all those things which can mimic it. So now we have to figure out, okay, in GCP, in Azure, what is it equivalent of that? And then we realize, okay, Kafka can be found in every other cloud, like Kafka like implementation. So, if we can find a one to one mapping, we just start using it, and then it's a simple wrapper around it for the rest of the ecosystem. So that's first what we try to do. Second thing is, for example, something which is very specific to AWS, like s three.

S3 is their own offering. And definitely you have block storage in other cloud, but s three is specific to them. It's not like open source like Kafka, which is available for all cloud offerings. So, for that we have our own microservice which wraps over it, which to the rest of the system gives all the guarantees of s three about the SLAs, about a consistency or all those SLAs, it gives or replication everything. Right. And we know that when we are connecting to another cloud, if that cloud is not offering those guarantees, our microservice internally will handle that so that the rest of the system can still operate on those guarantees. And about consistency or application or SLAs. So that's what we are doing as of now.

Matt Brickey

Have you envisioned yet, your organization being the scale to need private infrastructure? And how would that play into your architectural decisions that you're making today?

Shikhar Agarwal

Yeah, so as of now, we have not reached a scale to think of a private infrastructure. As of now, I think we are still few years, at least a few years. And even after a few years, I'm not sure whether that would be a wise decision or not. But again, technology thing changes so much, so I never said never to anything. Yeah, but we are not there for private cloud as of now.

Matt Brickey

Kind of a similar question. Have you started looking at the amount of machine learning and AI that you're using? There's a certain inflection point where I think the AI and the machine learning will actually be more than your cloud infrastructure. Right. From a billing standpoint.

Shikhar Agarwal

Oh, absolutely. So that's another area that we are looking into. So as of now, I'll just tell you some of the use cases that might be interesting. So, some obvious one is just the customer support chat bot where what we have realized 80% of the queries can be automated, right? Because the queries come. Hey, can you confirm my TSA Precheck number is on my ticket or my vegetarian meal is on the file? Or can you give me three options from New York to San Francisco? Right? So, all these queries we can totally automate. Some other interesting things we are doing is when you search for airline like flights, the fare rules that come are not structured something like is it refundable or not? What is the penalty for changes that is there?

But hey, in case you do not show up to the thing, what is the penalty for that, right? Or if you cancel a flight, you get a credit. What are the rules for redeeming that credit? Are there any blackout days for that? All those are unstructured, so we are parsing it to make structure out of it. Similarly, when we get hotel content, the images come at hotel level, not at room level. So, you had one queen room, two queen room, one king room and one suite. But the images are coming at hotel level, and they are not tagged of to which room this image belongs. So now we are trying to do ML image recognition. Hey, in this room is it one bed or is it two bed or is it a queen bed or a king bed so that you can match these.

So, there are a lot of use cases of AI and ML in the travel industry, right? So, we are starting to do, or starting to some of it we are already doing like the chat bot thing and we are closing, monitoring the bill. But you're absolutely right that at some point of time the bill starts becoming very high and at that point in time we'll start thinking what model we need to build in house, because it's not like certain models can remain static. There's not a lot of changes that happen again and again, all those decisions. Then we'll have to figure out what models to bring in house versus what APIs to use for off-site.

Matt Brickey

You mentioned the use of datadog earlier, which I get from understanding and monitoring your infrastructure. Have you invested any time and energy into tools or strategies around the cost optimization or the utilization monitoring? I know you mentioned you had four or five different environments, things of that. It sounds like you have some visibility there, but I'm curious where your head's at.

Shikhar Agarwal

Yeah, absolutely. So, I might not know the exact tools that we are using, but our head of DevOps and productivity tools, he is monitoring all those things. I know that we reduce the cost by 30% of our AWS by switching the instances in a non-production environment to tiny like mini and small and all those things. So we did reduce a lot of cost there. But last I know, our team was exploring some tools that can optimize or can give us more visibility on where we can optimize and reduce the cost. And I also went to the AWS reinvent in Vegas that happened last year and also saw a lot of companies who are saying that they can help us reduce the data doc cost.

Like they have intermediate layers that are built where they need not send all the events to Datadog and some of them they can just do on their layer and send selective events to Datadog. So we have not explored those yet because our team is just fully on building feature set and for our customers. But at some point of time I foresee, I think in Q three is when, according to our roadmap, some critical feature that will be delivered. Q Three is where we'll start exploring in more details, at least on my plate. But our team might be exploring those. Sure.

Matt Brickey

Yeah, sure. As you grow, it'll get more and more important. So I'm curious about the people, specifically the ones you managing. How are you approaching hiring and really just maintaining your kind of startup culture as you get larger, right. And you scale and scale, you get further away from that startup bubble, right. It was so much fun. When you're talking about a handful of people working on it, how do you maintain that kind of spirit, if you will?

Shikhar Agarwal

Yeah. So I would say that maintaining that culture is really hard. It takes a lot of effort and lot of repetition of the same things again and again. Right? So a few things that we do is, first of all, we know that meeting in person is really important to build a social currency. Like, we have offices India, London, New York, Palo Alto. So what we do, like every year we get the entire company together at one event. And every quarter I go to India to meet the Bangalore team and then we go for and most of our engineering managers and PMS product manager going there. Then we do Bollywood dance parties. You'll just spend some time in the informal setting to do that. So that is one, to just build a social currency, right. Two other things that we do.

One is definitely hackathon, which is pretty usual that every year we do some hackathon to just again have that innovative or fast moving pace culture. Second thing that we do is, I've told every. So we have different pods, Air pod, hotel pod, and also I've told the leaders of every pod that every quarter you need to tell me or demo me something innovative in your area, some new tool that might have come, some new technology that might have come. It's not necessary that we will adopt it or we will buy it, but at least it will tell us what all innovation is happening in the industry, right? So at least that also makes sure that innovation is flowing through. Rest is just lot of good deadlines.

Like for example, even now we have raised good money, but we have less resources than the number of features we have to build. So just keeping those tight deadlines and measuring them effectively, having weekly update meetings or biweekly cabinet meetings, just to make sure that we are all working and very focused, so that DNA is just built so that is still there.

Matt Brickey

As you guys grow and scale, what are the things that keeps you up at night? From a business perspective, from a culture perspective, I'm just kind of curious where your head lies with regards to the future and some of the concerns that you're trying to overcome.

Shikhar Agarwal

Absolutely. So I think there are a lot of concerns. Very frank, I by nature very conservative when it comes to executing, because what I feel is that a lot of people, for example, 350 people, are dependent on us and are giving us their entire time, their profession time, and hoping that we'll help them build their career and help in their daily lives. So a few things on business side. One is that we have closed a lot of deals, as I was mentioning, a lot of fortune, ten companies and a lot of enterprises. How do we make sure, and most of them are to be launched in Q Two, Q three this year, right? And so for me, it's like if they are effectively launched on a platform, like one of it is already launched with 1000 users and they're rampant, 100,000 users very soon.

But just making sure that all these enterprises successfully launched, the site is up, it doesn't come down. People are able to just book their travel and actually travel. So that definitely something that is on top of my mind and why it's on the top of my mind because what I've noticed is that when you are in this high pace environment, when the velocity is really high, the quality tends to be great. There are a lot of bugs that are coming. There are a lot of regressions that are happening. People are pushing code without writing tests, right. We have APIs now. You introduce a new field or deprecate a field without thinking about if some customer is using those APIs, right? So if you really think about it, yes. The end result is customer being successfully launched.

But what keep me up is just the quality of the product. And so now what we are doing is we know our velocity and pace is really good, but now we're balancing it with quality, that it's okay that our velocity reduces a bit of pace, reduce a bit. But it has to be enterprise grade, well tested, intuitive to use product without any regressions and all. So that's what we are focusing on and we are building the right thing. So just that entire testing and bugs and quality part. And the second thing again is that how fast you can patch a system when your enterprise is down, which comes to entire deployment process and a CI CD process.

So the deployment, the hot fix of how fast you can do, that's the other thing that we are working on before we can launch all these enterprise in Q two, Q three. Those are some of the, you're in.

Matt Brickey

The unenviable position of being in a zero tolerance platform, right. It has to work all the time because with corporate travel it happens all the time literally around the globe.

35:17 - The Future of Travel Technology and Spotnana's Vision

Matt Brickey

So I'm curious, as you see the industry, sorry, as you see the industry innovate going forward, what things are you seeing that are emerging technology wise, business wise that could really transform the travel experience the next 510 15 years?

Shikhar Agarwal

Absolutely. So one thing that I didn't mention what we did because it was not in context. But now that you ask if so what we have done over the past two years is we started this spot on a company to disrupt the corporate travel world, right? But when COVID happened and all and other travel agencies were facing a lot of problems, we thought why don't we just develop a platform that anyone can use, a travel platform. So this is what we call travel as a service, TAS. So what we have done in Spotnana are two things. One is travel as a service. This is like an AWS platform, right? And just like Amazon uses AWS, our corporate travel solution uses our own platform as our first customer of their platform.

But this platform is now also powering like Brex is a company, is a fintech company, credit card and expense company, Brex. They last year launched Brex travel to its corporate clients. Brex Travel is built on top of Spotnana's TAS offering travel as a service offering, right? So now what we have seen is that as the industry is growing, one is definitely like coming especially for the travel, two things are happening. So one is we develop this platform. The second is instead of going through aggregators, they're coming like Amidair Sabre, we do go through them. But we are also directly connecting to the airlines, directly to United Airlines, directly to American Airlines and all, which again, has never been done before. Right. The reason I'm mentioning is to give you some context.

Now, what the industry is moving towards is all these airlines want to give you dynamic pricing. So Delta wants to know, hey, are you also looking for United? If that let me reduce the price. So all the dynamic pricing industry is realizing that lot of money is to be made on ancillaries like your Wi Fi priority boarding, baggage lounge passes. As of now, in existing things that you think, like all the travel tools when you're booking a flight, they don't give an option to buy WIFI or a lounge pass. So how can they integrate this ancillary shopping along with your normal shopping, what you're doing of the flights, that's a huge change that the industry again is moving towards. From travel perspective, from technology perspective, definitely AI is huge for the industry, right? So they're rapidly talking about adopting AI and ML.

So that is one thing. Blockchain is huge for the industry. There's lot of inefficiency as payments goes like, there are so many settlements and so many players. When you buy a united ticket on Expedia, there's so many players who have to take a piece of that price. So all that is pretty inefficient. So blockchain is one big thing where smart contracts can help. Another big thing is sustainability in the industry. So how do we make sure that if something can be done on trains and rail travel and not flights, it's more sustainable? A lot of those things are happening in the industry as of now. And definitely everyone talks about cloud and all those things. So that are some buzzwords that the industry is so backward that these things are new for them, which is like.

Matt Brickey

Been around a while. Yeah. It needed some disruption, I'd say. So I'm curious about. So let's say that I'm a user. Are you also working on things to integrate AI to make it easier for me just to verbally tell your platform what I want. Right. It seems that a lot of the, you mentioned chat bots earlier, but a lot of the interaction with humans seems to be just, well, just tell me what you want, right, and I'll translate kind of what you're looking for and give you some options. Do you see that happening in your industry as well?

Shikhar Agarwal

Yeah, so we definitely see that. In fact, one of our hackathons, which happened last year, this was one of the hackathon projects that someone integrated voice to text and then used our APIs. And not just that, it was also multilingual. So you can say in Hindi and French and use translation. So, we see that we have not productionized this, nor we know of a tool which is adopted to a large extent which supports this. But I feel that this can be big, especially when you are doing thinking about, again, you are just on your mobile, you're driving, or you're just traveling and you just want bot hey, I want to book this flight from this to this for my meeting, send me some options. Right.

So those are the things that I feel that I'm pretty confident that the world will move towards, and that is something that we are keeping on top of it, not productionize it. But yeah, that would be something exciting.

Matt Brickey

I could definitely see something like that, especially for international travel where you're not familiar with where you're going. I want to stay in a four-star restaurant or four star hotel. I need a bus pass, I need a train pass, all those kinds of things. You feed it the information and it provides you the options. I think that'd be very good. As an international traveler myself from time to time, to have somebody safe, you.

Shikhar Agarwal

Read them for me. Totally. And given the latest, like after chat GBT and all the LLMs that have come through. So, voice to text is pretty advanced now, but now even the LLM side of content production and understanding, it has become very robust. So, I feel that now it's the right time to actually leverage all these technologies.

Matt Brickey

So curious here, how do you keep AI on your side and not have it disrupt you?

Shikhar Agarwal

So just to make sure that I understand, are you talking more in terms of people losing job or in what context?

Matt Brickey

Sorry, no, more of them. I guess as AI becomes more robust and chat GBT can maybe do some more things, how do you make sure that you're utilizing it? As it grows as a platform and the language models learn more and hopefully don't get all whitewashed away, but they learn more and they get better and better. How do you make sure that you stay on top of that as well as the advances to your own platform?

Shikhar Agarwal

Yeah, so I think for that, my answer is very simple. Innovation has to be bottoms up, right? From my point of view, I have constant time, constant number of hours in a day. You can just read so many things or do so many things, right? So, if you just build a culture of innovation, because what I've observed are engineers know lot more than I do of what's coming out there, right. They talk to their friends, they read different newsletters and all those things. So as long as engineers have a voice and you build a culture of bottoms up innovation, these are things that you can definitely be on top of. And you can then after you know it from people, connect it to your business, either existing or new potential opportunities, and that's how you even fund it.

Matt Brickey

Good answer. Yeah, that's a good answer. A couple more right before we wrap up. I'm curious, what lessons have you learned so far working in the travel industry? Being that it's so heavily regulated, it varies so much between country and country. I'm just kind of curious what the highlights are there, the lessons that you've learned.

Shikhar Agarwal

Yeah. So that's a very interesting question. So, the lessons that I've learned is that to disrupt an industry, even if not disrupt, to become a big player in an industry which is so much regulated and it's so different from region to region, is to have experts in your team helping you. So that is number one. Right? So, for example, we hired one of the coos of Amex as part of our team. We hired an influencer as part of a partnership team. So, getting those industrial experts who have done that, been there, done that really helps because first is to get all the domain expertise. The second is they have connections, so they know people in airlines and hotels, they know people in settlement payment, which can really unblock it. Right?

Similarly, we hired one of the heads of a company which actually developed all the APIs of airlines like American airline to lead our air product. So, they also know how to unblock us. Right. So that is number one. Number two is just patience and that perseverance, that things will happen. Like you have to be patient just to get a license to operate French rail, SNCF and TGVs. It's a long process to get the French government give you a license to enable you to book travel on that. Right. So, you have to be persistent. You have to have right people on the ground. You have to have the route accounting team to fall legal entities there. So, lesson learned in general is, one, it is hard. You have to be patient.

Perseverance, that is number one, you have to have the gut, the money and resources and bandwidth to fight those battles, because these are battles. These are like operationally heavy battles, stressful battles. You have to do it. You have to have the right team. I cannot stress less, right team is needed because they open up lot of doors for you, especially in these regulated things. And fourth is just that attitude that you have to then do it right because it's hard. That's why a lot of people have not solved it. It's a deterrent, but if you solve it, then you become one of the few players who can really grow big. So, the opportunity is huge. That's how Uber did it. It fought all these taxi unions and all those things.

And that's how we take inspiration from these, and we have to do it what it is required.

Matt Brickey

So, Shikhar, our final question. At the beginning of our conversation, you mentioned Bill Gates was a bit of an inspiration for you. So, I'm curious, what do you hope that your legacy in technology ultimately is when this is all said and done?

Shikhar Agarwal

Yeah, that's something I always think about. And a lot of people ask, I think for me, and this is not ego, I really see the opportunity where we are going. I would love to be the travel phase, the technical phase of travel. That's the legacy. So that's personally for me, for Spotnana, I wanted to live beyond me and beyond my co-founder and beyond the thing, right? So, we want to create a generational company that can survive generations and become the central, like sort of the operating system of the travel industry. Because of our travel as a service platform, can it become the central piece on which an entire ecosystem revolves? Just like there's AWs, entire ecosystem revolves. So, can we become that centerpiece?

So, I think those are some of the aspirational goals that we have, and we'll see how much we can achieve. We'll see.

Matt Brickey

That's great. Shikhar, thank you for your time and best wishes to you and your organization.

Shikhar Agarwal
Thank you. Thanks a lot, Matt, for your time.