"The personal feedback was on being user centric. So, one of the things that I felt at was splitting the fare. So if you remember, you and your friend can take a ride and then after getting dropped off, split the fare between you two. So this was broken and nobody noticed it, and he was the one who read the reviews online, on the app store and then he figured it out. And, that was one of the feedback. Always keep on a lookout for how your product is doing. Connect your work to its impact, and how you can grow it. So that was one of the feedback I still remember." - Karthik Gandhi
Karthik Gandhi is an engineering leader at Stripe. His current role is to help Stripe expand into SEA and India. Before Stripe, Karthik worked at Grab, where he managed critical economic levers such as Pricing, Surge Engine, Forecasting, Demand Shaping, Supply Shaping, and the ML platform that powered it all.
Karthik built his career in Silicon Valley before moving to Singapore. He started his career at pre-IPO VMware as a kernel developer. Later he started an NLP startup (General Sentiment) with his advisor Prof. Steven Skiena. Even though the company didn't succeed, the experience shaped his career.
Karthik was inspired to become a leader at Microsoft, where he was part of Azure team. His most significant career breakthrough came when he joined Uber as one of the early employees and spent close to 4 years there. During his time at Uber, Karthik helped Uber build its business in China, built its core infrastructure that helped Uber scale to millions of transactions/second. Later, Karthik transitioned to build autonomous vehicles and trucks. Karthik became part of the Uber Freight team when the automated trucks program was suspended and helped build a marketplace from scratch.
Karthik graduated with a Master of Science from Stony Brook University, New York, and later did a distributed systems specialization with the University of Illinois, Urbana-Champaign. He moved to Singapore in June 2019, where he lives with his wife (Software Engineer@Facebook) and his 3-year-old daughter. He is into adventure sports such as ice climbing, diving, canyoning, skiing, and any outdoor activity depending on the season. His indoor activities include reading books about leadership and human evolution - past, present, and the future, tinkering with my Raspberry PI cluster for distributed systems development.
You can find our community discussions on the podcast episode at
Please forward this insight or invite friends at https://whatsapp.com/channel/0029VakR55X6BIElUEvkN02e
Jeremy Au: [00:00:00] Welcome to BRAVE. Be inspired by the best leaders of Southeast Asia tech. Build the future, learn from our past and stay human in between. I'm Jeremy Au, a VC, founder, and father. Join us for transcripts, analysis and community at www.jeremyau.com.
Hey, good to have you on board Karthik. I'm really excited to share your journey with the whole show.
Karthik Gandhi: [00:00:35] Thank you for the opportunity, Jeremy. Happy to be here.
Jeremy Au: [00:00:39] Well, you've had a tremendous journey ranging from obviously the States, to working at multiple brand name startups as they'll call it, right? In terms of the engineering leadership, and you've made it all the way to Southeast Asia, and along the way you've managed to add Travis Kalanick as one of your role models and mentors. So we'll touch along on all those stories. But for those who don't know you yet, tell us a little bit about Karthik. Tell us about your professional journey.
Karthik Gandhi: [00:01:07] Hello, everyone. Thanks for dialing in. My name is Karthik. Right now I am a engineering manager at Stripe, Singapore. My team works on expanding Stripe in the Southeast Asian market. If you don't know Stripe, Stripe is in 43 countries, and is able to accept payments in 109 different currencies. Yeah going back to my journey, I got into computer science during high school. My first computer was when I was in my ninth grade, and ever since I fell in love with programming. And I attended top coder contest, online programming contest that used to be my undergrad in high school. Since finishing my undergrad, I started contributing to Linux Kernel and major projects such as Hadoop Edgespace at the time.
Every part of my career, my interest, was picked by the technology, the advancements in technology. So I was always learning, always looking at what is happening in the industry to a point where this is an interesting story which I can share. Back in 2011, I bought 12 bitcoins and sold them almost immediately just to understand what it is, to be a cryptocurrency and what it means to be a blockchain. So, that's how I was driven by what's happening in the industry.
And later I was building distributor systems, so databases that can facilitate up to millions of transactions and to self-driving cars and trucks, but the motion I was building, the motion planning stack and currently I am looking at Fintech. So as you can see, always technology that has paved my career, or shaped my career.
Jeremy Au: [00:02:39] Awesome. Well, I'm sure we'll get to chat little bit more about Uber Freight and all the various crazy things you've done at multiple levels of technology, difficulty levels I'll say. I want to kind of like go back to the beginning a little bit. So, tell us more. So, you did Master of Science in computer science and you're doing applied algorithms. Tell us more, what was it like to be researching applied algorithms back then?
Karthik Gandhi: [00:03:04] I worked with Professor Steven Skiena for those... He was my advisor for those who don't know Professor Steven Skiena. He's one of the best known figures and applied algorithms. He has written a lot of books and his books are very popular among the online coding community. So, it was a dream came true when I joined his lab. Yeah, researching was more about learning from him, it was mostly what he was doing. And then, here's an interesting thing. Skiena doesn't tell you what to do, he gives you directions. He was one of my role models, we'll talk about the role models later but he was one of my role models.
And the way works is that you show something, he'll give you a small problem statement and then you go work on it, and then show it to him and be like, "okay can we improve this?" And then you go back and work on it, can we improve this. And, then finally he'll be able to piece together all the work you've done in to a paper, right, so that is how applied algorithms work.
And for those who are from computer science background, the difference between applied and theoretical is that applied means essentially you are taking an industry problem and working on it. For example, flight planning. Now, can you go from point A to point B with four different stops and still make money, right, and not have a terrible customer experience. So that's the kind of a practical problem you take and work on it. In my case, the practical problem that we worked on was sentiment analysis.
At that time, I was 28, 2008, 2009. At that time, the challenge was that we are getting a lot of data from Twitter and Facebook, they were coming up and then onward BlockSpace was also becoming very popular. Now, we wanted to analyze how these impact the way people think about politicians and celebrities in general. So, we developed an algorithm that could tell you that given a personality like Barack Obama, it can tell you whether his sentiment in the public perception is positive or not based on the actions he took previous day or the next day. So, it's essentially crunching a large amount of data and coming up, bringing up these insights. So, that was the research that we were doing at the time.
Jeremy Au: [00:05:07] So, you're doing research and obviously there's a big pile full of folks at that point of time, right, which is do they either go into more academia, right, and go the research route or to join the... I don't know, the forces of Mordor or the evil private sector, right? So, tell us more of like did you know going as a master's student, did you already know that academia wasn't there for you or were you debating it?
Karthik Gandhi: [00:05:35] I'm glad that you asked this question. I wanted to be a PhD student actually. I did complete my research proficiency exams, but what happened was that as part of the research that we were doing, I was pretty excited and I went to a couple of conferences and then presented the results of the analysis we had made, it was just a grad student who was very excited, just pressing the results. And then in one of the conferences, I believe it was 2008, Hadoop conference, couple of VCs approached me and said, "Look, this sounds like a very good idea. Can we commercialize it?" I was like, "Yeah, why not?" Right? If somebody's interested, why not do this? And that's where we started pivoting from a research project to more of a commercial project.
The company was called General Sentiment. We founded it, at the time it was coming up, we cannot use lab resources for commercial purposes so we had to move all the number crunching to data crunching to AWS and there was a desk and then four researchers working with VCs. It was an interesting experience I would say. And we had to build essentially everything from scratch, and two years in I realized that I am more of a engineer than a researcher at that time. And, it was pretty clear to me that going back wasn't really an option which is where I decided to just continue being in the industry instead of going back to academia. But my original intent was to be in academia.
Jeremy Au: [00:07:06] That's very interesting. I never knew about that. So, these [VCs] were like the snake in the garden of Eden, right, asking you to eat a forbidden fruit of industry, right?
Karthik Gandhi: [00:07:20] Yeah. Yeah, when somebody approaches you and says they'll give you money and I've never seen that much in my life, I was like, "Yeah great, sounds good."
Jeremy Au: [00:07:32] That's a great story. So wait, so tell us more. Okay, so your... Bring us to that room. So you're presenting that paper and then some VC walked up to you and gave you the business card, took you out for a meal. Like, what was it like?
Karthik Gandhi: [00:07:46] To my advisor, it was mostly on the side to be really honest. So, it was my advisor working closely with the investors and then founding that company. I was mostly listening to the conversations about how there is potential for making a lot of money with this kind of data, how NBA players would be interested in it and then even media or Hollywood celebrities would be interested in this kind of data. It was novel. Like, data analysis is now something that we take for granted but at that time took processing something like a two terabyte of data was not commonplace.
Frankly, there was a time it struck me. Look, I was just doing it for fun and somebody thinks that there is a lot of money to be made, so why not, right? That was what was going on in my mind when these conversations were going on.
Jeremy Au: [00:08:34] Yeah, I mean it's actually a common story, right? Like, advisors get, you know, they're building something that's really interesting and then there's a interested age of commercializing and some people go back to academia well after the commercialization successful or less successful. What was it like at your... I guess that was your first startup right? I guess watching the commercialization of this technology and transition from academia to commercialization.
Karthik Gandhi: [00:08:58] I worked 20-hour days. I remember that vividly. I'd come home, I'd sleep, wake up around 5:00 AM and just go 5:00 AM to 6:00 AM and then go back to work and eat at the office. It was just complete dedication, I ignored personal relationships to a large extent. That is one part I vividly remember. The other part was showing progress, right? Before that you're in academia and you're kind of working with your professor, there's no goal, it was mostly set by your own self, and then now we have a boss, we have a CEO who is tracking progress and wants to know what's going on. So, that was also a little different.
And, the third one is budget management and team management hiring. So, before that you were just working with a bunch PhD researchers with you. Now we need to go hire people, more importantly, identify what kind of people you want, so that was also a big change I remember. And then... These are positive things.
So lastly, one of the things that I remember that didn't work was outsourcing prematurely. So at that time, outsourcing was a big deal. For this kind of sentiment data, you can see there's two parts. One is the back-end part, the data crunching part where you build all the models and data points and then there's a front-end part where you plot them on a graph. One of the ideas that came from the VCs was that okay let's just outsource this front-end part to one of the Asian countries, that way we can save costs, right? Turned out it wasn't a great idea after all because it was premature. And we had a... I vividly remember this. We had a deadline for launch and the previous day we found that the data that was being sent by back-end and the way the data was plotted was completely wrong. I remember just closing my laptop, walking out and car and driving to Boston, and did not look at my phone, did not look at my laptop for four days and just waited there.
Jeremy Au: [00:10:55] Wait, wait, wait. You were so frustrated that you closed your laptop for four days? What? Wait, how does that work?
Karthik Gandhi: [00:11:06] Okay, tomorrow's the launch. Everything looked wrong. What do I do? The clear answer is not going to happen, we need to plan it. But at least, you know, I worked 20-hour days, let me take a break so that I can come back fresh and start again. So that was the intention behind it.
Jeremy Au: [00:11:26] That was your first ever job, right? How old were you? What year was that?
Karthik Gandhi: [00:11:29] This happened in 2009, so I was 23.
Jeremy Au: [00:11:32] Yeah, so pretty much your first job, your first startup job. Wow, that's a very common first startup job experience, but it's a tough one to be in every single time, right?
Karthik Gandhi: [00:11:45] Actually it wasn't first job because when I graduated from undergrad, I joined VMware . That was my first job. I was there briefly and that's where I realized my undergrad education can only take me so much and there are way too many smart people, and that's what led me to go to grad school and that's what led me to my own startup, so yeah.
Jeremy Au: [00:12:06] Oh, awesome. Yeah, and then how did that startup journey wrap up? So, you were there for a few years. You eventually transitioned to working at Juniper or at Microsoft, yeah?
Karthik Gandhi: [00:12:19] Sure. Couple of things happened as we were working on this product, as we were building it. At some point it was starting to look like we don't need specialists. It was more of okay let's just save costs, let's try to sell what we already have. We need operators who can run the job daily and someone to maintain the front-end and the back-end, and not make any changes. And it started to... To me, it looked like it is going into maintenance more and that's where I decided that maybe it's time for me to move on. And, there were also personal commitments that I had to make at the time which means I had to find a real job and make a lot of money. So, those are the two points where I decided to leave startup and then move to Silicon Valley and start over again.
Jeremy Au: [00:13:14] So tell us about this third job. So, you had been VMware which was in 2007 which was a long, long time ago where everyone runs... everything runs off VMware these days anyway. And then, you were at this startup for two years and then after that you went to the Bay Area. Tell us more.
Karthik Gandhi: [00:13:32] Yes, so Bay Area was refreshing I would say. So, I was completely burned out from my experience at that startup working 20 hours days, it wasn't sustainable in the long-term. So, Bay Area in a way is where I recovered and started seeing the world very differently. And when I say Bay Area, I was living in San Francisco with San Franciscans and they're a rare breed. It looks like a lot of them were priced in the last seven years, but before that there were a lot of amazing people in that city who helped me to become a person who I am today. So, I owe a lot to that city in a way, my favorite city.
So there, Juniper Networks, I chose that company because it aligned with my previous expertise that was building kernel and networking and file systems. So I thought, you know what, I'm just going to take this job. All my personal issues, effectively grow in my career and also recover from my burnout. So those are the criteria that I used to join Juniper. And, Juniper operates very differently. They ship boxes, they build these network boxes that everybody, all the big companies, you know, Google and Microsoft they use in their data centers. And these network boxes can essentially move all the bits and pieces across the network. So, the box need to be built perfectly before it gets shipped to the customer.
So, they follow a traditional software development model. What that means is that there is a planning phase and there is a development phase, there's a extensive testing phase. So, it was not to compare to Facebook or the online web services, you have to build, deploy, if it breaks roll back, build, deploy. So that wasn't the cycle these traditional companies follow.
It kind of gave me some level of freedom to travel and work from anywhere. So, even though the work from anywhere was popular in the last one year, I was doing it at that time. I traveled, I don't even remember the number of countries. But I would finish my design, get it approved and then travel to a country of my choice, work from there, come back briefly, show my work and then go back again, work from some other country. That is what I was doing at the time and was incredibly rewarding experience.
Jeremy Au: [00:15:43] Yeah, I just want to say these are like huge brand name companies today, right? Because I think like VMware is like, what, 30,000 employees now, you know? I think like Juniper's maybe like 10,000 employees, so you know like... You were like pretty early... I wouldn't say obviously that you weren't an early employee, but these were fast growth leaders back in the day as well.
And, also what's interesting is that at this point in time within like less than 10 years I would say, or I think closer to five years you had basically been growing up in India. Then you were in Illinois, Illinois in Chicago, and then after that you were in New York. And now you're in the Bay Area, right? That's like what, seven years in all these different places. So tell us more a little bit about that before we talk about your Grab, your Uber, your Microsoft experience. You've done these moves, right? So, what was it like to be, I guess, a fresh grad from India going in to the States?
Karthik Gandhi: [00:16:40] I think I mentioned this during my VMware experience. I was lucky enough to be employed there, I would say. It wasn't imposter syndrome which I'll talk about later, but I still feel I was lucky enough to be employed there because that's where I met incredible people. VMware used to be a family company. The CEO was the wife of the chief scientist was Mendel Rosenblum, the professor of Stanford, and then pretty much all the chief scientists were his PhD students at the time.
So, I worked on specifically CPU scheduling and memory management problems which put me in touch with some of these researchers. And, that is what made me realize that there is a world is out there and there are incredibly smart people out there. And some of them became my mentor, some of them gave me advice on how to progress from VMware to different parts of my career and that's what led me to a choice of moving to US. So, I hope that clarifies how I chose US.
And then after choosing US, applying for universities, I required some guidance and some of the employees at VMware helped me pick the universities, applied there. And one of the criterias was cost because coming from India, education, especially Master's education in the US is really expensive. So, I ended up writing to professors where this is my top score, I've done this work in the past, this is all public information. Can you waive the fees or can you help with my interests? Again, Professor Steven Skiena gracefully agreed to do that and that was what led me to land in New York and join Stony Brook.
And then from there to Chicago was mostly for research purposes. I wanted to extend my research in to distributed systems, so it was more like okay let me go there. There is Professor Indranil Gupta who is really good at it, so let me go there and learn from him. So, that was the journey there.
And Bay Area was like a dream place. I never thought about it until I was burned out, until I was blank. Now when I say burnout, you wake up and you're like, "What do I do?", or "How is my day going to be? I just want to lie down." And New York, it snows and it's cold for six months, so you just want to lie down and sleep. And one of my roommates identified that, "Look, you're not well. You need to get help", and that's when I decided I need a change of scene. And that's where I'd randomly chose Bay... It wasn't a conscious choice, I randomly chose Bay Area. I randomly chose San Francisco to be my next destination.
Jeremy Au: [00:19:09] Yeah, I definitely love San Francisco and the Bay Area as well. I actually had this song called... It's by I believe it's Train and it's like San Francisco. Is this the title? Is this called San Francisco? And every time I play it, listen to it I always like... bring a smile to my face. Anyway, so what was it about San Francisco that made it magical to you?
Karthik Gandhi: [00:19:28] Certainly the people and the culture, right? So I can share couple of examples of San Francisco back in 2010 for people to relate. I used to love biking. It's a sunny city. So there was an incident where I got a flat tire. I was just waiting there, didn't know what to do. And then a person was kind of standing next to me. I didn't even ask her for help and she came to me and said, "Look, I just phoned my boyfriend. He lives nearby, he's going to come and he has all the tools to fix your bike." I'm like, "Great, this is fantastic."
I think there was one night I was working late and around 10:30 I went to Panera Bread in the neighborhood and it was closed. And, like okay, looks like I'm not going to get any dinner and then I started walking back. And then an old lady came running, she was like, "Okay, okay, let me open. I just closed it. Seems like you're hungry so here you go. I'm not going to charge you for it." So these are couple of instances I vividly remember, but then there were so many instances where people were so friendly, people inclusive, made me feel like part of the community. And that is what was magical about San Francisco at the time.
Jeremy Au: [00:20:33] Yeah, I remember actually roughly in the same time frame. I was in the Bay because I was studying at UC Berkeley in Berkeley and there was a very similar magic going on. And, I think still today as well where I think the people are just I think so friendly to people, right, and immigrants and just the kindness, and I don't know, the hippy-ness. You know, the free spirit there was just so much fun and such a breath of fresh air.
Karthik Gandhi: [00:21:04] I want to add one point about the hippy-ness. So, maybe Jeremy you must have met them too. I'm proud to say that I met the real hippies, not the hippies with ample accessories all over their body. So those are the latest editions of the pretend hippies so don't fall for that, anyone who's listening to it. So the real hippies are very nice and they're kind of outliers. And that's another thing I really loved about San Francisco, there everybody questioned anything that I believe in my life until that moment in time in a nice way. So, the conversation would go like, "Oh look, why the drink and smoking is allowed but if you take marijuana it's illegal?" I'm not vouching for drugs or anything, but just a curious question. Why are drinks widely available but not the one? You know, they can have the same effect. So these are kind of questions they ask me every day.
Why do you believe in God? Why do you believe... So everything I've known in my life was challenged there and that is what made me think who I am, what I believe in and what I should be believing in. So, that is another part I really love.
Jeremy Au: [00:22:09] Yeah, in the same vain of like gratitude towards San Francisco, beyond of course my tie dye shirts which I have a couple of them, which I always wear at home and my wife just kind of like, you know? I think she's gotten used to it by now, but she used to do a bit of a eye roll every time I wear my tie dye T-shirts at home. And I always go out to like functions and I'll be like, "Can I wear this?", and then she's like, "No you can't. You're respectable now." And I'm like, "No, why? I can still wear my tie dye shirt." Sometimes I manage to get her permission, right.
But I think it's just so fun to... I think Berkeley is, you know, they're very proud of the free speech movement and their role during the Vietnam war and so on and so forth . I think there was a huge amount of... obviously there's a lot of mythology around it. But I think it was nice to hear such a intellectual diversity of opinion where people felt like they could just share what they felt. And, I think I disagreed with a lot of it as long as I'm sure a lot of people disagreed with whatever was being said. But I think UC Berkeley worked very hard to preserve that open questioning culture as much as it could and I really benefited from it. And I think it had permeated over the whole Bay Area . That is a interesting part, right?
So you spend three years at Juniper growing rapidly and so forth, and then you're like let's move to Seattle, right? So you love the Bay Area but not enough to say, "You know what, if I grew up in India, I might as well go all the way tour all of America, right?" So Chicago, the East Coast, SF and now Seattle, right? So, tell us more about that move.
Karthik Gandhi: [00:23:40] It's for personal reasons. So my wife, she went to city of Washington so it made sense for me to be there with her supporting her. That was the decision behind moving there. And also with Microsoft, what I felt was that it was a natural extension to what I was doing before. I was getting really curious about this called cloud computing and how it works. I've used AWS in the past during my startup days, but then Microsoft was just getting started and I thought it would be good learning opportunity to see how it is built from scratch . And the choice of not picking Amazon was because I felt the product was already mature. So that is why I picked Microsoft and that's why I picked Seattle.
Jeremy Au: [00:24:22] Okay, I'm going to have two questions. One about work and I got to ask the first question that comes out of my mind. How did you meet your wife?
Karthik Gandhi: [00:24:28] Online. It's very...
Jeremy Au: [00:24:33] Online in the States?
Karthik Gandhi: [00:24:34] Yeah.
Jeremy Au: [00:24:35] Okay. So wait, that was a very early. How did you meet her online? Was there like a bulletin board or was it like an intermediary helped out or what?
Karthik Gandhi: [00:24:44] Oh, no no. So still, I think match.com and eharmony was there. It's not-
Jeremy Au: [00:24:50] Was it match.com or eharmony? Which one was it?
Karthik Gandhi: [00:24:52] I think eharmony.
Jeremy Au: [00:24:54] eharmony, okay. I'm disappointed because my mentor, Peng, at Monk's Hill was co-founder for match.com, so he has a lot of connections under his belt. eharmony is good as well, that's good. You know, a lot of people think eharmony and match.com are very old, but I've met a lot of couples, married couples. I think there's a different use case for Tinder, I think. Tinder is not the same marriage case I'll say.
Karthik Gandhi: [00:25:20] Oh, no no. I've never signed up for Tinder or any of that. I don't know. I cannot say lucky or unlucky because I don't know how it works, but yeah.
Jeremy Au: [00:25:30] Yeah. Sometimes my wife pulls up Tinder just to show that she can get lots of matches. And she's like, "This shows that I have value in the stock market", and I'm like... She's like... And I'm like, "Okay, I get it. Yes." It is what it is. This is interesting way to demonstrate market value.
Karthik Gandhi: [00:25:49] Instead of swiping up, you're swiping left and right I guess. It's still a good pastime.
Jeremy Au: [00:25:53] Yeah. So, yeah that's my wife everybody, yeah. And that's interesting, right? Because what I'm noticing here is each time around you just got bored of mature technology. You started... It was like, got it from the early emerging side, you brought it to the mature side and then once it became mature, you're like, "Great, let's build data centers for Microsoft", and that was in 2013, 2015 where just... Again, pretty much the... Obviously it wasn't founded that time, but there was a big part of the rapid growth and development and the depth of databases at Microsoft.
Karthik Gandhi: [00:26:31] For those of you who don't know, I started at Microsoft when Steve Ballmer was still CEO and the transition happened to Satya Nadella and I think I can say this now. At the time it was still running one [inaudible] machine per machine. In other words, there was one-to-one mapping and they were trying to rapidly scale. My task was that I was building a networking equipment for these big data centers. Now, seeing the big picture on how these data centers should be built will give you a perspective on how these data centers look like. And everyone should know it because your data's going through this and your data is being stored under these data centers. There is one in Singapore as well for Microsoft, by the way.
So, let's talk about the Columbia data center. Columbia is a city that is about four hours away from Seattle that has one of the biggest data centers for Microsoft. We're talking about millions of servers, I'm not even kidding. We're talking about millions of servers connected to each other and there are different topologies. There is cloud topology, [inaudible] topology, these are not randomly wired up together. That's not how it works. There are different well thought of topologies under which these machines are deployed in the data centers and power and cooling are two major issues.
So, that reason for Columbia is that it's naturally cool and dry so the cost of cooling is fairly low. And when it comes to power, they have their dedicated generators to powering millions of machines. In the event of failure from these power grids, there was a battery backup. It's like fuel cells, the traditional lithium-ion battery. In the event of the failure of batteries, there are steam engines, locomotive engines that could immediately start and power up these machines. So, it was a massive set up. We take these for granted. We take the phone swipe left and right or up and down. There's a lot of machinery that goes behind that. That is how complex building data center is.
Jeremy Au: [00:28:22] At this point of time you've kind of like gone through a couple of different culture changes of different companies. Around those two years, you were seeing a transition between Ballmer and Satya, right? And obviously we now know the market for foremen and Satya has been quite tremendous, right, both in innovation, approach. Did you see any of that culture change or strategy change? Was it obvious then that Satya was going to be a great CEO? Or was it more like oh no we're not really... we can't really tell whether there's a big change or not?
Karthik Gandhi: [00:28:55] What was obvious at the time was that a change was necessary. But what wasn't obvious was if there was going to be the changes that the company requires, at least at the initial stages. But to me one of the personal experiences with Satya was that I had the fortune of meeting him once and I asked him, "I know you're Satya, but there is a perception of Microsoft that I want to change. You know, if I go to Bay Area, everybody's about Apple. Apple is a cool company, Steve Jobs is a great CEO and a great product manager. Whereas, if you look at developing countries in other markets, Microsoft is still very popular. And, I still love all the Microsoft products I'm using. What is causing this perception difference between Apple and Microsoft?" And he said, "Perception goes in cycles, right. So, at some point a company will be very popular, and at some point some other company will be very popular. These cycles are not something we should focus on. We should focus on building quality products."
So you see, he immediately shifted the strategy from everything on Windows to everything everywhere. So, it was the biggest strategy change which was very unpopular, but he was one of the leaders who made that change when I was there. What that means is that everybody was moving towards Apple or Android phones, but then they were starting to also use Google products in those... Google or Apple products in those phones. So Satya Nadella said, "Look, let's build an Office for Apple. Let's build Office for Android phones." So, that is how he started looking in to the proliferation of platforms and how he wants to be in those platforms. I hope that made sense though. Earlier the strategy was let's keep everything Windows... Office exclusive to Windows, but later he changed, started to say we have to be in these devices.
Jeremy Au: [00:30:43] So, we're going to assume everybody knows what happened. So, we're just going to jump straight in to more like the inside out, right, which is why was that change unpopular?
Karthik Gandhi: [00:30:52] Partly because Microsoft completed a $7 billion acquisition of Nokia devices at that time if you remember. So, it was to build phones and tablets and surfaces which we still use. The shift was kind of a move away from a decision that was made like six months to eight months ago, right. And also, Microsoft enjoyed a dominance in industry for a very long time, so accepting that okay now Apple is a platform that we need to cater to, Android is a platform that we need to cater to was a mental shift for employees of Microsoft. Those are the reasons what made the change very difficult.
Jeremy Au: [00:31:31] And, you did your two years at Microsoft so you already kind of made the whole rounds of all the giants of Silicon Valley or recent technology. And then you decided to join Uber as an engineering leader, right? So, tell us more about why you joined. What was Uber like in those early days when you joined?
Karthik Gandhi: [00:31:49] Sure. So, this is where the story gets really exciting. So, Microsoft was one of the biggest companies I've worked for, it's 100,000 employees. It's a big machinery and it was still early in my career. So, what I felt was that the execution was still well thought out, but it was kind of slow for what I was looking for and that is what made me look for a new job. And even before my wife graduated and joined a job, I told her, "Listen, I'm leaving. You know, you can come and join me later." So, in a way I left and moved back to Bay Area just because I was missing the energy of the Bay Area.
And at that time, I considered a few companies. I think Quora, Snapchat, Uber. Those were the three companies that were up and coming good people that I know from the past. And the choice of Uber was primarily because of the team members and the vision.
Jeremy Au: [00:32:46] So tell us more. At the time, how big was the Uber engineering team when you joined?
Karthik Gandhi: [00:32:51] I think there was 300 something. That's how early it was at the time and Uber was growing crazily. Looking back, I still feel I was very luck to be part of that juggernaut .
Jeremy Au: [00:33:04] Yeah. When you said 300, 300 employees or 300 engineers?
Karthik Gandhi: [00:33:07] 300 engineers.
Jeremy Au: [00:33:08] Awesome. And so, obviously at 300 engineers it's an interesting place because in order of magnitude, smaller than Microsoft, VMware and Juniper. So, you must have been doing a lot of like organizational change, processes change, engineering culture change. What was that like coming in?
Karthik Gandhi: [00:33:27] Yes. So... And this is where my leadership journey kind of vividly getting started. So, earlier when I had my startup, there was certain level of leadership but it was at Uber where the pace at which the company was growing and the number of people we had building things did not match up. Which means you're wearing a lot of hats and there is no proper structure, no proper organization, everybody does everything. People are reporting there are no formal managers looking at career planning. No formal VPs of engineering looking at organization structures.
So, I kind of ended up doing quite a lot at the initial 10 to 12 months. That was the time I started realizing about how best to set this organization and I worked with some of the senior leaders and helped them look in to... look at Uber's infrastructure organization that I was part of at the time, and how to split that in to better structures.
Jeremy Au: [00:34:23] So at this time, you had shared with me previously that you were happy to have gotten to know Travis and to have been coached by him on this. Tell us a little bit more about what was it like meeting Travis because none of us have never met him, we've only read about him. So, tell us more. What was it like?
Karthik Gandhi: [00:34:39] Sure. So, I know Travis has been bit of a controversial figure in the media, but then one thing we should remember is that he did build a massive industry from scratch. We all owe that to him. And, there are three things I want to mention during this podcast. One is how to build a mission driven organization. Travis was at the front and center of everything that is Uber. The brand, the employees and the mission, he was front and center there. And during the 2017... I don't know if everyone knows what happened. People did not leave in rows, they stayed back. Everybody looked in to the company and said, "What have we built?" How to fix it rather than the taking flight. And that is in some ways attributed to Travis and his leadership. People loved the company a lot and they were giving 150%.
So, Travis is a leader that definitely built a company, a mission driven company. So when I say mission driven company, it's also important to talk about what was the mission of Uber from looking at it. It's very easy to dismiss it as a taxi company. But if you look very closely about the experience of getting a taxi before uber days, again Jeremy you can relate to it, in San Francisco you have to literally jump in front of a taxi for him to stop and even consider you as a passenger during a Friday night.
So, why was taxis very scarce? It was because of a concept called medallions. Taxi medallions are very scarce. It's a rare resource that lets you drive a taxi. It's kind of a license that you need to have before you can drive a taxi. And if the medallions were fewer in number, artificially fewer in number and they go up to $500,000 at $1,000,000, that is completely unaffordable for regular taxi drivers. So taxi mafias, they hold these things and they will have a proportional number of cars which they rent out in a hourly basis or four hour basis, which is why you meet a lot of grumpy drivers. Because within those few hours, they need to make as much money as possible before returning their car. So, that was how the whole industry was set up. And then, Travis was set to change that.
And this wasn't only the case with US, it was the case with many different countries. So that's what he was set to change. It was obviously very difficult and when he shared this personal journey on how to we can fix it, that's what motivated the company. And then, there was another learning from this which was holding unpopular opinion is okay at times. If you're a CEO of a company like Uber, you cannot please everyone. The journey is treacherous so unpopular opinions are going to be there.
But one of the unpopular opinions I remember was tips. It was a controversial topic. Again, this is very specific to the US context. If you go to a restaurant, a 15% tip is kind of mandatory. Why is that? It's because back in the day, there is a $7.25 minimum wage that was set, but then it doesn't apply to restaurant employees because their tip is factored into that $7.25. So essentially, something that was supposed to be given for a good service became part of their base minimum wage. So Travis was against that and he said, "Look, let's not have tip concept in our app at all because it will factor into the earnings of the drivers which may not be good for them in the long-term." But, you know how it all went. It makes sense now, but...
The other one is looking at root cause rather than symptoms. Free lunch is something that everybody takes for granted. But one thing we needed to remember is that if you don't like it, you can go out and eat. So there was a all-hands meeting where somebody walked up to Travis and said, "The lunch yesterday was horrible. Why don't we change it?" His response was, "Look, maybe you should be fired or your manager should be fired." I was like, what are you saying? And he said, "If among all the things that we're doing, the mission, the work, the challenges that we have, if you have a problem with something like lunch, maybe because you're not happy on a day to day basis. And if you're not happy because you don't like the work or you're not happy because your manager is not doing a good job. So, that's what we should be addressing rather than the food itself." So those are three examples I want to give.
Jeremy Au: [00:38:38] Interesting. And you also shared that he gave you... At that time, you were also choosing to build up Uber Freight, right? Which was at that time felt kind of crazy when it was first announced in the public. So what was it like doing that, building out a new initiative?
Karthik Gandhi: [00:38:53] So, he was kind of on his way out. It happened suddenly but it was part of the self-driving truck initiative. Uber acquired Otto and Anthony Levandowski joined it as part of this acquisition at the time. That is what kick started the truck initiative, the Uber truck initiative. The idea was that building self-driving trucks is much easier than building self-driving cars. Partly because trucks travels through the highway a long time, whereas cars need to travel through cities a lot. So the original idea was that a driver, somebody would pick up all the packages and then he'll drop off the packages 50 miles away from the city. And then an autonomous truck will pick this up, travel through the highway and then drop it off 50 miles before the destination city, from which point a manual driver will pick it up. That was the idea and unfortunately it got suspended.
The team was motivated and so they pitched the idea saying, "Look, we thought about it as a freight business, right? Why don't we just do freight business and instead of having trucks, let's just have... I mean self-driving trucks, we can just have like truck drivers in our network. We have a brand already so let's just do it." And it was in between Travis's transition so it was part him and then part the new executive team that looked into this idea before approving it.
Jeremy Au: [00:40:13] And you also shared that before that Travis had also given you direct feedback, so could you share about that?
Karthik Gandhi: [00:40:19] Yeah, that was mostly to him outages, right? He's a busy man, so he has other things to do. So, whenever there was an outage like surges not working or driver software or allocation is not working, that's where he was very involved. So, the personal feedback was on being that user centric. So, one of the things that I felt at the time failed, that I failed, was splitting the fare. So if you remember, you and your friend can take a ride and then after getting dropped off, split the fare between you two. So this was broken and nobody noticed it, and he was the one who read the reviews online, on the app store and then he figured it out. And, that was one of the feedback. Always, always keep on a lookout for how your product is doing. Connect your work to the impact of the work and how you can grow the impact. So that was one of the feedback I still remember.
Jeremy Au: [00:41:13] Okay, and so you've been doing this Uber Freight, you've been managing the whole network, building Cassandra as a service and all these other frameworks. And then, you decided to join Argo AI. So, tell us more about that.
Karthik Gandhi: [00:41:26] So, Argo I would still feel that it was an extension that I was doing back at Uber ATG. Uber ATG for those of you don't know, it's called Uber Advanced Technologies Group and that's where I was working at, you know, self-driving cars and trucks. And, Uber Freight was later part of that too.
So Argo was much an extension. So if you look at the self-driving car industry, it's very small. So there are two major research groups, one from MIT and the other one from CMU. And Uber ATG was primarily found by researchers from CMU, the Carnegie Mellon University. So some of them felt that it's not just enough for us to build a software, but we need to integrate very closely with the manufacturers and build a self-driving form, a car from ground up. It doesn't make sense to just put something on the top of the vehicle and make it a self-driving vehicle. So, there was a motivation behind that. And I knew this works before so it was more like okay I've been at Uber for some time and this is an opportunity to build a remote office and a remote team which I've never done before. Well, I've done partly but not like something from scratch.
I took this as an opportunity to build a remote team and also contribute what I've learnt of building self-driving vehicles at Uber. That is how I ended up at Argo. Just known folks, a family environment, yeah.
Jeremy Au: [00:42:44] Okay, yeah. Okay, so this all makes sense. You know, you loved San Francisco, you doubled down in Uber and all the self-driving. And then suddenly, Singapore, right? So, how does that work? How do you move from the Bay Area which you love to Southeast Asia?
Karthik Gandhi: [00:43:03] Interesting question. So, it's life choices, right. So the reason why I moved to Seattle. So the reason for moving to Singapore is based on my parents. They were the primary reason why I moved to Singapore. The original idea was to move back to India, they live in India. But then the move to India would necessitate that I live in either Bangalore or Hyderabad which are like metro cities where I can find really good companies and positions. And, going to my hometown from these cities would take longer than going to my hometown from Singapore which is very interesting, so that's where the choice of Singapore comes in.
And also, I realized that the transition from Bay Area would be much simpler and I would accept the transition if the transition was smoother. And at that point, Singapore made sense to me.
Jeremy Au: [00:43:50] Did you grow up in Tamil Nadu?
Karthik Gandhi: [00:43:52] Yep, totally.
Jeremy Au: [00:43:54] Yeah, I meet lots of people from Tamil Nadu. Well it's like a direct line or boat, I guess, from Tamil Nadu to Singapore, right? So all my friends... North Indian friends are like this stuff is all Tamil Nadu, right, in Singapore, right?
Karthik Gandhi: [00:44:11] Well, it's one of the official languages.
Jeremy Au: [00:44:13] Exactly. So yeah, it's interesting that Tamil Nadu is closer to Singapore in many ways than Hyderabad.
Karthik Gandhi: [00:44:16] It's funny.
Jeremy Au: [00:44:25] So, Grab obviously it seems at least to me it feels like if it was extrapolated, it would be straightforward because the recruiter was probably like, "Oh my god, I'm so excited you worked at Uber. Of course we'll have you on board." Is that how the conversation went?
Karthik Gandhi: [00:44:37] Thought about it a lot and I think it wasn't very easy moving to Singapore to be honest. Jeremy, the way you described it was what I was thinking, but it wasn't very easy. The reason why it wasn't very easy is that in the Bay Area, if you were looking for a job you just go out there and within couple of weeks, you'll have 10 offers waiting for you. That's not how Singapore works and Singapore market is very limited, and the scale of the company was also very limited.
I narrowed down to few companies, ones like Indeed... Stripe at the time surprisingly and I'll come to that, and then Grab. So, originally I was supposed to be the head of grocery business and it later got shut down. So, I got an offer, I decided to move and then it was shut down. And, I only quit my previous job and it was in the middle of that and it was a difficult period . Luckily, some other team on Marketplace, they reached out and said, "Look, we have a new opportunity, not like the scope that you were looking for, but can you take it?" And, I just took it.
And the reason why I didn't go with Stripe or Indeed is that at the time, after building remote office for Argo, I realized that it is a challenge. Being in a remote office... it's like being time warp. You work in certain ways and then the headquarters is kind of moving really fast, they're different. So, you don't have all the information and then everyday you sync up and you get new level of information, but then you're not part of the day to day discussions. So, that is what drove me to pick a company that has headquarters in Singapore and at that time it was just Grab.
Jeremy Au: [00:46:15] Interesting, yeah. And, at this point of time obviously you've worked in a lot the titans of Silicon Valley, then the new emerging titans called Uber in SF and then obviously now at Grab. I'm just kind of curious, were there any big culture differences that you saw?
Karthik Gandhi: [00:46:33] Oh yeah. It was a big change for me. I was new to the market and as I mentioned, the reason why I picked the company with the headquarters is because I would get to learn from them. It's the heart of the business, right? So a lot of processes are set up, a lot of development is set up. That was the reason. But then again, coming back to the culture differences, what I found moving to Singapore was that a lot of my management experience had to be revisited.
I was used to these debates, right? So you have a team to manage and then there is project that needs to be done and there is a lot of ideas. You also come up with an idea, your team members come up with an idea. You debate it up and then you come up with a good solution. I come here and then I'm like in the first meeting, "Hey, here's an idea", and everybody says, "Yes, let's do it." Like, no no no. That's not how it works. We need to think about it and then look at the other solutions and pick the best one. So that was one of the first differences in terms of culture.
The other one is large meetings, really, really large meetings. So, you have so many people and you have this meeting and there were so many stakeholders to agree to a particular decision. It wasn't something that I was used to before as well. Meetings happen in a small group, decisions made in small groups just to make everybody to be efficient, and the organization circle structure supports that. But that wasn't the case here.
There was another difference. The third one is it's the ownership aspect, right? So I think we should talk about this a little bit later in detail, but the ownership aspect was missing which was also surprising. In other words, you're building a product, something like surge pricing, something like forecasting system but then you don't really connect with the results. Somebody told you to do it, you're just doing it but you don't really connect to the user who would be using that product. Whereas in Silicon Valley, the connection is really obvious and every engineer wants to know how this product is going to be used by a user, and they work very closely with it. So here that was another change I need to get used to as well. So those are three main things I remember.
Jeremy Au: [00:48:37] No, that's interesting. And after that, now you're at Stripe.
Karthik Gandhi: [00:48:42] Yeah. After my experience at Grab I realized that when it was up and coming, that was fine. I think I got a lot of good opportunities, good team members. I was able to make several changes partly to address all the problems that I discovered when I moved here. So, I changed my approach first. One was from saying, "Look we should do this" to "Can we do this?" More of a questioning, inquisitive approach and that helped with the first problem.
The second problem is, again, having extensive moments discussing asynchronously rather than having large meetings. And then the third one is I'd ran a hackathon for my entire team where they traveled to different parts of Grab's markets, interacted with the customers, came back with ideas and we ran a hackathon. So those are kind of things that Grab enabled me to do and I'm really thankful for that, for those opportunities.
And later on, last year, Grab was really hit by the pandemic. That's where we went from scaling everywhere to like okay let's focus on our core business and how do we become profitable. These are which are good goals, but then I thought I could be better... For my career's sake, I thought that it would be good to go to a company that is scaling and expanding during these times and that's where Stripe came in. It was an easy choice, I just went back and said, "Look, we interviewed back in 2019." So it's just like, "Can I come in now?" And Stripe was like, "Okay, you know, fine".
Jeremy Au: [00:50:07] One thing that's over this career as you used to be an individual contributor, as an engineer and you started leading small teams then larger teams. At Grab, you're managing 30 software engineers and data scientists. So how do you feel like your leadership style had to change as an engineering leader over time?
Karthik Gandhi: [00:50:26] One thing I stopped caring about was the titles. I started caring more about the scope than the title itself. So what is the scope of my role? How can I make impact? So, that was something that changed over with the period of my leadership. Earlier I used to think that bigger teams are better, but now I don't hold that belief. So smaller, nimble teams make faster progress, create more impact than larger teams because the coordination costs itself is much, much higher than the actual product. So, that is one belief I remember.
The other one is technical complementing with the strategic view of the business, is one of the rarest skills that I need to develop. Initially, delegating work was very difficult. We've mentioned technical challenges, what I get excited by. So when I started managing teams, delegating work was one of the challenging things. Now I learned that it is one of the essential skills and I'm able to do it more effectively.
And then, the other one it's a lot of patience, assertiveness and an exceptional ability to listen. So, earlier I used to jump around when somebody says, "Look, I have this problem", I'm like, "Okay. Wait, let me go solve it." Now I listen. I go to the other team, listen to their side of the story, get every single side of the story and then make a wise decision that is agreeable to both. So, it's not just about the team that I manage, it's about the company as well. So that is another learning that I had.
And then, flexing the management style. Engineering is a creative profession, right? So, traditional approach such as look this is the way, this is my management style, it doesn't really work. You have to cater to individuals that you're managing. So that also took some time to realize. Some people are independent, they want to be independent. You just give them a one line statement, they'll run with it, they'll finish it. But for some people, they require a lot of guidance, hand holding which is okay, but you need to flex your style according to each of your team members. That's also another thing that changed over the career.
Jeremy Au: [00:52:21] Wrapping things up here, if you were to go back 10 years, where were you at that point of time and what advice would you give yourself back then?
Karthik Gandhi: [00:52:31] I would still say keep learning and keep learning and keep experimenting. That's what got me where I am and I should never ever stop that. So, going back I would just tell them like always, always, always keep trying. Always learn new things. Improve yourself everyday. And yeah, that's life.
Jeremy Au: [00:52:49] Awesome. Thanks so much Karthik, I really appreciate it. So moving on to the Q and A section, so if anybody wants just to raise their hand feel free to raise your hand and we'll be able to have Karthik answer any questions that you have. If you raise your hand, about his time at Stripe, at Grab, at Uber, all these other places. So if you raise your hand, we'll go from there.
Awesome. So Karthik, how can people reach out to you if they want to hear more or reach out to you for more questions?
Karthik Gandhi: [00:53:16] Sure. LinkedIn is the best way. free to connect with me, I'm happy to answer questions there. And also, Jeremy has an awesome site too. Feel free to become a member and then we can talk there as well.
Jeremy Au: [00:53:26] Awesome. We now have a new question. So Lee, feel free to ask your question.
Lee: [00:53:30] Hey Karthik. Can you talk a little bit about the engineering culture at Stripe and compare it to what you experienced at Grab and Uber?
Karthik Gandhi: [00:53:39] Good question. Stripe is known for their engineering culture. In general, it is highly positive culture. So, the way I can compare Stripe to Uber, I would say Uber is more of a better comparison, is that Uber grew more rapidly and partly because it's a consuming facing company whereas Stripe is a B2B company. And the way planning works at Uber is more bottom-up, meaning you can hire any number of people and we've mentioned there was about 300 employees, within a year we had about 5000 employees. So that's how fast Uber grew and they could be working on many different things. Even teams we're working on same things, but in different ways.
Stripe is more measured in their growth. So, when it comes to hiring, when it comes to expanding, there is a lot more thoughtful approach to what we need to do than saying look you have to conquer all the markets and let's just keep trying things. So the experimentation part is very low. The other aspect of Uber versus Stripe is that Stripe is a highly regulated business. You're moving fast and breaking things and experimenting doesn't really work, so we need to take a measured approach. Okay, does this work and this is acceptable to the regulators, so we need to keep all that in mind.
The third part is the security, right? Security and privacy are paramount when it comes to financial transactions. So, the focus on that would mean that there are lot of reviews and how to deploy, how to develop products, so that's also something different at Stripe compared to Uber.
Lee: [00:55:16] Okay. That makes sense, you have to do much more. Methodical and controlled in how you go about it. Cool, thanks. That's interesting.
Jeremy Au: [00:55:22] Anyone else out there would like to raise a question? Or Lee if you have a second question feel free to ask.
Lee: [00:55:27] Okay, I have another question. How does Stripe split its functionality between its Singapore operations versus those in San Francisco office?
Karthik Gandhi: [00:55:38] So, it's not really I would say is a separation. Yes, there is a focus on Southeast Asian or in general APAC markets. But then we collaborate very closely with the SF teams in terms of day to day developments. The only difference I would say is that there is like a lot more platform work in San Francisco versus a lot more methodic work in Singapore.
Lee: [00:56:03] I see, that makes sense. So you basically use the platform to build out the products for this market?
Karthik Gandhi: [00:56:09] Yeah.
Lee: [00:56:10] Thanks, that was helpful, thank you.
Jeremy Au: [00:56:13] Karthik, I actually I just wanted to ask a question based on what everybody was just asking about. So, one of the big questions that I often get from SEA turtles for example, you know, returning to Southeast Asia about... Just yesterday I was just having a call with someone who's thinking about moving in from Seattle to Singapore and she has a partner and everything. And, you've done a lot of moves with a partner. So how do you have that conversation with your spouse about the career moves? Is that a tough one? Do you have to like have a long dinner or something like that? How does it work?
Karthik Gandhi: [00:56:50] So, she's actually listening by the way, so let me be measured in terms of my response. The interesting part is not to show up one day and say, "Look, we are moving to Singapore" or "Look, we are moving to some other country", right? So that doesn't work. And, my partner actually works, she's also a software engineer. She's a software engineer at Facebook today and she was a software engineer at Amazon bay Area. So, it wasn't really an easy move. In fact, if I have to relate back to my job search in Singapore, I would say that my partner found it more difficult to find a job that has the scope, the salary and the responsibilities that she was looking for at that time.
It was in some ways a more difficult move for her than me moving into a leadership role. But the key is giving a heads up. The move to Singapore, a move to anywhere outside of Bay Area was a one year conversation for us. Because after we had our first child we were thinking what's best for her. Is it best for her to be close to grandparents? It was a constant discussion and it was an everyday discussion. So, I think that's key. And every day, there was a concern that's coming up and then we address it, then the next day, new concern comes up we address it. So, you get the point, right?
But just before the move itself, there was a miscommunication. Yes, everything was smooth at that point, but there was a miscommunication about when I would be moving to Singapore and she was caught unplanned. I said, "Okay, in two months I'm moving. I signed the offer", and she's like, "Wait, wait, wait. We were discussing and suddenly you're moving? I haven't even started my job preparation yet." And I was like, "Oh, I told you before." "No, no, no we were discussing. You never told that you've made a decision."
So that was a mess on my part too which means that she had to scramble at the last minute and manage the child all by herself for couple of months. At the same time, attending job interviews which made it very, very difficult. I remember the times where she'll have the child, which was one and a half years at the time, in her hand while attending interviews for Singapore companies in the middle of the night. So, yeah it could have been better. Glad that it worked out.
Jeremy Au: [00:58:58] Yeah, I think my wife and I also in our move back to Southeast Asia, it was like a two year conversation and it wasn't easy. A lot of miscommunication along the way. I always tell people we may never pull the trigger if it wasn't for COVID because I was like... In February, March and I was reading news and the English news was pretty chill. It was like, oh you know, China is working on it. I was reading the news in Chinese and they go panicking and bricking their neighbors in, right? And I was like, there's a bit of a disconnect here a little bit, right, between...
I remember my American friends were like, "Wow, they're building a hospital in a week. Very interesting", and I was like, "Okay, it's not". You're like wait, it's interesting I think but they're not building it for fun. Not for a Guinness record. They're like desperate, right? So I think as I remember I convinced my wife I said, "It's probably going to come. I have a thesis that it's not going to be great in New York where we are." We had a nice Times Square apartment. But 40% of New York doesn't have health insurance. It does not going to go well. Why don't we just do like a... I pitched to her like as a two week holiday in Singapore.
Karthik Gandhi: [01:00:10] That was interesting. Yes, I should've done that too. That's a good tip.
Jeremy Au: [01:00:15] That wasn't a good tip.
Karthik Gandhi: [01:00:16] So, in my case it was my first trip. The move was her first trip to Singapore which wasn't the way it should be. Yes, that's another good tip Jeremy.
Jeremy Au: [01:00:26] I wouldn't recommend that. I think that was very one off situation. And even then, I remember she literally packed for two weeks and she almost didn't make the move with me, I actually need to be honest. The night before she was like, "I'm just going to stay here. I don't think it's going to get that bad." Because I think it was like less than 10 cases and I was a little bit more skeptical so I packed for like four weeks, so. Yeah, but that's also... We ended up making a move and for the record, yeah that's how I also ended up accidentally... during that move that's how I ended up with pandemic baby.
So things get busy when you're moving, right, and you just make a mistake and woops. And that's how I end up having a three month baby today, you know?
Karthik Gandhi: [01:01:13] Oh, congratulations. We should give them a new name instead of pandemic babies, that doesn't sound very appealing. Like Millennials, Gen Zs, there should be a new name for them.
Jeremy Au: [01:01:21] Yeah, I always like to call them miracle babies which now I know it just means everybody says like "we screwed up."
Karthik Gandhi: [01:01:29] But, in a way it's a little blessing to be in Singapore because you can go to the hospital, have a baby instead of being worried about any of the COVID situation.
Jeremy Au: [01:01:39] Yeah, for sure. I think that's what we were discovering while the situation was getting rapidly worse in the States, and life was getting better and then we discovered pretty late in the process, we're like, "Oh, we have a kid" and we're like, "Okay." Well, that changes. Now we had to reset our conversation again about how to move your spouse, right, so that was another fun conversation. I'll say unfun to be honest, to be utterly honest, but it worked out. Everything worked out, yeah.
Karthik Gandhi: [01:02:07] One thing I definitely want to say is that in terms of child care, in terms of supporting working mothers, I think Singapore has done a great job.
Jeremy Au: [01:02:15] Yeah, I think that's something that's really underappreciated for sure. It's just the ease of child care in this situation is just dramatically better, especially also of the year of COVID, right?
I just want to welcome Bruce. I think Bruce has a question. So if anybody has any questions, feel free to raise your hand and then I'll be able to bring you up the stage. Yeah, Bruce?
Bruce: [01:02:31] Thank you for setting up the room. I'm a fintech entrepreneur based in Shenzhen, China. I used to work in Seattle at Microsoft 10 years ago before I moved back to China. So, my question is I would like to hear Karthik's perspective in terms of the engineering culture since you moved from US to Singapore. I want to hear your perspective, like US and the Asia or most specifically Southeast Asia or maybe you...
Also, Stripe has engineering team in India ?
Karthik Gandhi: [01:03:04] Oh yeah, Stripe does have a team in India where we are hiring and in Singapore obviously we're hiring too. So, I think the Stripe culture is more uniform, so maybe that's not very interesting to cover. So probably go back to Grab and how differences are there . Yeah, so I covered one of the aspects which was ownership which was very different for me coming from the Bay Area. So the engineers are connect to mostly... The companies that I worked with, the engineers are connected to the mission, or anyone who's building the product was connected to the mission. They're independent, would just run with it and there was a notion of more startup vision ownership within in the company.
But then here what I found is that decision made top-down primarily. So it's like somebody decides, somebody approves, all the decisions, there is a hierarchy mentality. That was one of the changes coming from the Bay Area to Singapore.
The other one which I covered earlier was as a result of having this hierarchy organization and meetings, sorry, approval process, there are a lot of meetings where a large number of people come in and it can drag for several days before a particular decision is made. So, the decision-making process was very different.
The third one is in terms of leadership itself, the change was again is a product of being hierarchical is that the decision is not made bottom-up or even the planning is not made bottom-up. It's like everybody feels that whatever they're working on was decided by somebody else. So it's like oh I'm just doing what I was told, right? So that was very different for me coming from there as well.
Bruce: [01:04:40] Yeah, I think that it's my experience... Yeah, and what about the talent, in terms of the engineering talent? Is it easier to recruit in Singapore compared to the Bay Area?
Karthik Gandhi: [01:04:53] That is one of the challenges I'm facing right now and there are two aspects to it. One is the existing talent. So, I would say that Singapore has incredible talent and I'm speaking from my heart. The only thing that was missing was the direction. That was my experience so far in Singapore. They're really amazing. The education system here is rigorous enough to build solid, quality engineers on par with any US university, but then they lack the direction.
Today as I mentioned, you tell them, "Look, you need to work on this", and they're just like okay somebody decided I'm just going to work on it and then submit this project, right? Instead, if you give them the opportunity to think for themselves and own what they build, I saw incredible results. Yeah, so I would say the existing talent is there.
In terms of new talents, yes acquiring new talent is not as easy as it was in Bay Area and that's something I haven't' figured out yet. The way it used to work in the Bay Area is, okay you tap in to a network and say, "Look I'm looking for this engineer who knows X, Y, Z", and then the next day you have 10 candidates, and then a week later somebody clears the interview. And two weeks later, they're in your team. That's how fast the whole process was. Here companies... It's not uncommon to hear that there is a notice period of two months, so you try to hire and they come up with, "Oh but I cannot join for two months", and that's a challenge.
And the other part it's in terms of exposure, I find that even though they have talent, a lot of them were not exposed to different styles of working that could really help their career. So, I always face a dilemma of do I take a risk, hire them and coach them or should I just wait for someone who has done this before and hire that person. So those are kind of the challenges I'm facing today in terms of hiring.
Bruce: [01:06:35] Got it, got it. Thank you.
Jeremy Au: [01:06:36] Gaurav, go ahead with your question. And if anybody else has a question, feel free to raise your hand.
Guarav: [01:06:45] Sure, thanks a lot Jeremy. I have a question for Karthik. I come from Microsoft, Karthik, so we share the employer. Okay, so Karthik I just want to understand... I belong to product so I'm not in core engineer, but Microsoft is a big engineer place so I understand engineering well. So, what is the drastic difference working in India if you ever worked in India?
And Indian startups are doing pretty well, so and about Bay Area. So, what's the difference you feel? I would not compare it with the [inaudible] example because Singapore I think primarily similar, much more aggressive maybe than Indian companies, but culturally we are kind of aligned. But Bay Area might be very different because they are leading the world in terms of startups, so. What is the difference you felt as an Asian or as an Indian?
Karthik Gandhi: [01:07:36] Let me understand the question. So you're trying to understand the difference between the Indian startup ecosystem versus the Bay Area startup ecosystem? For me as an Asian, right?
Guarav: [01:07:46] Yes, correct.
Karthik Gandhi: [01:07:48] Yeah, so I briefly worked at VMware India before moving to US, so I have some context there. It could be really outdated because it was back in 2007, so I'm not going to call myself an expert in that area unfortunately. Yeah, sorry about that.
Guarav: [01:08:04] No problems, no problem. That's okay, fine.
Jeremy Au: [01:08:06] Guarav, do you have another follow up question rather than that? If not, we'll move to [Zidan] then.
Guarav: [01:08:12] No, Jeremy.
Jeremy Au: [01:08:13] Okay, Zidan go ahead and ask a question for Karthik.
Zidan: [01:08:16] Hey Jeremy, thanks for adding me in. Karthik, so I had a question. You just said that one of the challenges is whether you are ready to hire talent and then train them or hire those who are already trained in them. So, what's your opinion in it? Because some of my background, I'm a recent graduate, so obviously my perspective will be like I would love to be hired and trained so that I can get up to the mark. But as per your leadership experience, what do you feel is the better push for you?
Karthik Gandhi: [01:08:46] Yes, thanks for the question. The way I think about the new grad, and there is one who recently joined the team which is [inaudible] in the room, is that I need to set you up for success first. So, I'm not just looking out for the company, I'm looking out for your success as well. So, how can I make you successful? It is training, right? And if they're happy to trainee, I need to have the resources or people to train you. And when it comes to that, I look at the proportion of the team. How many senior people do I have and how is their time being utilized? Is there a room for me to get in a new grad and make them work with the senior talent so that they can learn and grow.
So, the way I make the decision is looking at some of the work you've done during grad school, your past internships and general knack for technology, right? One other thing I faced, the challenge I faced in Asia is that there is a set path for many. When I say that, some people think engineering or medicine is just the way to go, right? They may be interested, they may not be interested, but it's just like the way the society thinks. Let's just go with engineering or medicine. So, that's where I find some challenges in terms of new hires. They're not really motivated, it's very difficult to get them in and train them to be better software engineers.
Zidan: [01:10:09] Got it, got it. And what do you think? Was your experience good [inaudible] getting in new skilled... getting skilled software engineers versus training those who are out of school. Who do you think have performed better in your experience?
Karthik Gandhi: [01:10:26] I've had good experiences and bad experiences. So, I would say that in the Bay Area especially, the new grads have come in from different kinds of school, have different expectations and different attitudes. Some graduates they think they've done everything and there's just like nothing much to learn. They're hard to coach. Some people have more open-mindedness. They're like, "Okay, I'm willing to learn. I've done this, but I'm willing to learn", so they're more easy to adapt to the company culture and coach them to be better software engineers. So, I've had this experience with both. So having an open mind and saying that, "Look, I've done this but I'm willing to learn this", would really help in terms of training.
Zidan: [01:11:06] Got it, I think I really agree with you. Got it. Thanks Karthik.
Karthik Gandhi: [01:11:10] Sure.
Jeremy Au: [01:11:13] [Jaemin] , you'll be the last question.
Jaemin: [01:11:15] Yeah, hi Karthik. Thanks a lot for sharing. So, I'm also in engineering. I spent my time around Bay Area and down in New York as well. I'm also considering making a similar move back, closer to home, back to Southeast Asia and maybe Singapore. Just wondering if you have any specific advice on how to approach the planning, the [inaudible] process, eventually the transition for someone with a similar background.
Karthik Gandhi: [01:11:38] Mm-hmm [affirmative]. Networking. That's the key here. So you need to network with top leaders of the industries that you're interested in. In my case, I'd share that with Grab it was a little easier because there were a lot of ex-Uber colleagues there at Grab, so I just reached out to them and say, "Look, I'm looking for a move. Can you suggest teams?", and that's how I started the whole process. I would recommend the same to you. Plan for six months at least and don't take the first job that comes your way.
So, make sure you have done you analysis, make sure you've understood both sides the professional and the personal sides of living in Singapore. Which areas you want to live in, if you have kids, if you have a partner, working partner. See they are well supported in addition to you. But yeah, going back to the original point, networking is key. And then, understanding kind of work that is being done here is also key. That's where you should be starting first.
Jaemin: [01:12:31] Gotcha, thank you.
Karthik Gandhi: [01:12:33] And I suppose you have looked in to the visas and other situations as well, so that's also key especially during the COVID travel.
Jaemin: [01:12:41] Maybe one follow up question in terms of a lot of that you've been mentioning was around like the engineering culture [inaudible] expectation or like a mindset shift. Would you say that's something that... I know that it's been... like the things might have changed recently since there's been a lot of new innovation initiative in Singapore environment to like basically [inaudible] sector. But, do you any specific advice there?
Karthik Gandhi: [01:13:09] Let me understand your question. So you're asking for how the culture is changing in terms of technical companies and what they do and the government initiative supporting them?
Jaemin: [01:13:21] Yeah, exactly.
Karthik Gandhi: [01:13:22] Got it. Okay, yes. Singapore government has been very supportive of local initiatives, local hiring, local training, having part in university courses. And even quantum computing, whichas Jeremy know, the government has recently incentivized to build Singapore in to a tech hub. So the ecosystem exist and wherever there's a gap, the government is definitely stepping in to fix that.
Jaemin: [01:13:48] Gotcha, yeah. Thank you.
Jeremy Au: [01:13:50] Awesome, thank you so much everybody. Karthik, thank you so much for sharing the time that you have had today.
Karthik Gandhi: [01:13:55] Thank you Jeremy, thanks for having me.
Jeremy Au: [01:13:58] Thank you for listening to BRAVE. If you enjoyed this podcast, please share this episode with friends and colleagues. Sign up at www.jeremyau.com to discuss this episode with other community members in our forum. Stay well and stay brave.
Produced by Tan Yong Quan