CoRecursive: Coding Stories - Story: Software World Tour

Episode Date: November 2, 2022

Today story is from Son Luong Ngoc who shares what’s it was like for him to work and live in many different countries around the world, including working for AliBaba at the Xixi campus in Hangzhou, China. It’s a story of a software developer finding a place that fits them, a place that suits them. Episode Page Support The Show Subscribe To The Podcast Join The Newsletter

Hi, this is Co-Recursive and I'm Adam Gordon-Bell. Do you know the song, The Safety Dance by Men Without Hats? It's like, do-do-do-do-do-do-do-do. Early in my career, the second developer job that I had, it was in 2008, I worked at a web agency. It was a smallish office and they played music in the office over speakers. Mainly, I think, to drown out the people on the phone, support people, salespeople. The owner picked the station. And I think it was beamed right from 1982 because it was all 80s pop hits all the time.

And I know this isn't true, but in my mind, when I remember it, it was 50% or more The Safety Dance by Men Without Hats. I still have no idea what that song is about. But I left that job at some point and I went to Opertel. It was a software product company. And the first day I noticed how quiet it was. It was so nice. The web agency was culturally like a sales place. But Opertel, it was quirky and it was nerdy. I could tell right away. I just fit in better there. But maybe I wouldn't have known it without the safety dance, without hearing that song so much.

That's what today's story is about. It's not a story of a piece of software being built. It's a story of a software developer finding a place that fits them, a place that suits them. It's also going to be a bit of a world tour where we spend some time in various countries, including Canada and China. And I don't want to spoil too much, but elsewhere. So can you tell me who you are and what you do? Yeah, my name is Sun Lung. I contribute to various different open source projects, including Bazel. I have some contribution into Git and GitLab itself.

And yeah, that's a short summary. The earliest glimmer of Sun's career started with a lie. He was in high school at a friend's house, and the friend showed him a game. I think the name of it was Bounce or something, where you're a ball, and you were just trying to get from one platform to another, and eating candy or something like that, getting points. His friend claimed that he knew how to make games and that he could make this game. He was completely BSing by the way. Like he cannot make that. You need a graphic team with images to build that game. He alone cannot build that.

So Son tried to make his own game. And I went home and I just keep remembering like, how the hell do I make anything work with this Visual Basic program? Trying to drag the boxes and drag the button. This button is not the game that I'm looking to build. I can click it and create like a hello world. I can click it and create a text game where you answer yes or no and then choose your path. But that's not cool. And so Sun started taking some night classes in computer programming.

Where they teach you how to make computer program like web server and stuff. And I remember the teacher, they was like, man, you're in grade 10. Why are you here? My classmates were all final year university students and people who already have a job and trying to learn more. I was the only high school student there. The class was fun, but making a game still seemed like a mystery. So when Sun finished high school, because he was good at math and liked the idea of building a game, he went to the University of Waterloo in Waterloo, Ontario in Canada as a foreign student.

Waterloo was a culture shock. Sitting next to me is one Asian guy who was born and raised in Canada. And he was like using Vim and Latex to take notes. And I was like, what the hell is Latex? What the hell is Vim? And sitting on the left-hand side of is an indian guy who's taking a double degree program in both science and pure math the guy end up being a doctor and i remember a conversation on first year he was like oh did you buy this stock and i was like what you're you're

just turning 18 dude like how the hell do you know anything about stuff turns out his parent taught him like all that stuff and i remember like on third year i met a couple who was like after the final exam was going to the library to borrow math books to to go back and study after the final exam what the hell are you studying for and i was like man that entire environment was way too competitive coming from asia i was not expecting people to be that active learning in asia was something very passive whatever you are given to by the teacher you learn that and and that's it but in north amer and I think in Europe, if you were to learn something, then you must have a passion for it because you have a choice of choosing to learn that topic.

So once you have a passion for it, it means that you enjoy learning about it and you would spend a lot of time self-studying about it. So I did not have any of that. Even though I love math, I didn't love it that much. I just love it simply because I was good at it. This was Sun's first lifestyle adjustment, and there's going to be more. But finding his place at Waterloo was a challenge. It's a very serious school, but Sun found a way to make it work. I remember I was having a really interesting conversation about like, hey, what do you think gravity is? And people was actually like replying to you like, okay, here are like the theoretical physics about gravity. The university taught me two things. One is obviously like the computer science stuff is quite cool. But what it taught me is like,

I need to value this environment. When I enter a competitive environment, I should not be competitive, but I should befriend all these people. Because it turns out later in life, when I look back, I cannot find the same quality of people, the same smartness, the same speed that I was able to find in Waterloo. There's a path from being a foreign student in Canada to becoming a permanent resident. And that was Sun's goal. But all his friends who are Canadian seem to be getting jobs in the US, at Google or at Facebook or at wherever. In Canada, we call this the brain drain.

And maybe with remote work, this will slow down a little bit. But for Son, with his friends gone, he felt trapped in Waterloo and in Canada. So I was like getting really depressed and I was like studying straight for eight years without seeing my family at all. So no video calls, no air flight, nothing like that. So I decided, okay, I need a mental reset. And this mental reset means that I should go back to Vietnam and try to find something. After the quiet life of a math department in a small city in Canada, being back in Vietnam seemed like the future. A lot of expats, French, American, Russian,

everybody was moving to Vietnam or another Southeast Asian country because they like life over there. The weather is wonderful, beaches. If you ever had a chance to take a vacation, I definitely strongly recommend Southeast Asia. During this time, at least in his head, Son was still keeping his game dev dreams alive. I began in Vietnam starting to think like, hey, maybe I can open up after like one or two years of working there, I can collect enough money to open up a game company. To build his bankroll, he started a job as a business analyst at CSE.

They were making insurance-type software for various different companies. And turns out, back in 2016, a lot of insurance-type companies were still running on COBOL. Turning that into Java is hugely profitable. And yeah, we were having to deal with various different insurance companies. At the insurance company, Sun started to see a problem with his game-making plan. When you get into insurance, you start learning the value of reliability, the value of being able to assure something. People pay serious money for that.

And I was like, why do all these people who are obviously very smart, able to afford a lot of money to buy these insurance packages. Why do they value these so much? As a Waterloo math grad, Sun easily had the skills for understanding insurance and actuarial math. And it all started to make sense to him. There's some value here. You don't need to gamble everything. But there are percentage and there are value in safety net. You can actually quantitatively calculate that using an actuarial model behind all that insurance. And then all that math from Guadalupe start kicking in.

And I was like, oh, this all makes sense now. Okay, I can actually model out my life and my income and all that in a much better way than just like gambling everything away. And through this new lens, Sun's plan to start a game dev company, it just didn't seem to make sense. I was like, oh shit, that is actually gambling. And there's a lot of risk into all that. And I should not be pursuing that if I want to have, you know, a sustainable life, a healthy life. So Sun starts learning more software development, but he has a different perspective now because it's not about gaming.

It's just about building things and learning. Maybe being a developer can be his calling if he just throws himself into it. So I had a lot of time to study and I was like learning AWS, all that good stuff. And I landed a job in Lazada onto a team called 911. And you can guess what that team do. Emergency support for the tech platform. So Lazada is an e-commerce company in Southeast Asia. It's either a number one or number two at the time.

I think they were number one back in 2017. And they were selling in six different countries. Population-wise, Southeast Asia is huge and diverse. It includes Vietnam and Thailand and Laos, but also Singapore and the Philippines, Malaysia, Indonesia, and more. It's got 700,000 people. And each of those countries, people speak a different language or have different customs. Lazada was serving that market out of their main office in Vietnam.

A few hundred people worked there. I think at least 200 technical. And then we have the business side as well. So I remember we had at least three floors on a building. And I think each floor was carrying around like 100 people. You get free flow drinks. You have a cappuccino machine. You don't need to work long hours

unless you don't want to. What's it like to work in Vietnam? Very competitive, very cutthroat competitive. And I think it's not solely for Vietnam, but it's just like that for Asian countries in general. The tech stack is PHP. Monolith was being converted to microservice Golang running on Kubernetes.

Back in 2017, this is very modern, moving really fast. We have an R&D center in Moscow. We have an R&D center in Vietnam. And a data engineering center, I think, either in Thailand or Singapore or both, we have three different data centers that we manage by ourselves. The 911 team, well, it had a clear job. Receiving escalation from business side.

So because we have business entity in six different countries, including Hong Kong is seven. So we were receiving escalation from business entity. And then we escalate that over to tech side. And the job there was like to speed up the troubleshooting. So instead of 911 having to forward the ticket to the development team, then 911 just fix it. Just go there, find what's wrong, fix it immediately.

So I had my hands on all over the system. And yeah, I was like learning to be what now they would call an SRE, before Google released the definition of SRE. But yeah, like with like the background in insurance, I was immediately able to recognize why do a business was valuing this, because they want that stability. They want that reliability. And yeah, it's just a very interesting way for my previous job to somehow, you know, met the criteria of my new job. And yeah, it was very fun. It was a very fun time. For an e-commerce chain in Southeast Asia, the biggest day of the year isn't Black Friday. It's something else. Oh yeah. So back in Southeast Asia, folks in

North America and the European country don't know this, but Southeast Asia is heavily influenced by China. And China has big sales on Double 11 Day. So on the 11th of November, is the biggest sales in China. It's similar to Black Friday in North America or Boxing Day in Canada. So Double 11th is also the biggest sales in Lazada at the time. We opened up sales at midnight for all the countries. And before that sales coming in, we had to do, you know, stress testing for a month before that. And we were able to prepare for the load but on the big day a problem occurred so midnight of the 10th of november to the morning

of you know the 11th of november we detect the problem like 30 or 20 minutes in we were like oh we had a lot of order in malaysia. Wait, why everything stuck in pending payment? Wait a minute. Something is wrong here. Go back and check. Oh, it turns out everything was like being invalidated by payment processor. Turns out on the night, our credit card processor, a third party credit card processor in Malaysia just went down.

They just completely given up. All of the transaction we had sent over there was like completely disappearing. When you build up an e-commerce platform, you cannot build everything. You cannot build the e-logistics side. You cannot build the payment. You cannot build the notification part. So you need to use third-party. And when third-party integration goes wrong, you need to have a backup

plan for it. At the time, we don't have one from Malaysia. We had a war room going on in the 911 room and everybody was there. The payment team, the CEO, the CTO, everybody was in the room. And I had to crowd up the logs to see what's wrong. The CEO and CTO was like looking at me like, hey, look at the log and see what's wrong. And turns out it was payment. All of this was way before event-driven microservice. We had to like, you know, repass text log or JSON log into like some sort of pseudo CSV file. And then we have to manually data fix the database and then replay all that state turns out if you turn off all the cron job that invalidate the order that was not paid for one hours every other cron job also die in the system so yeah the system back then was like not at the best state

but that that's how we kept all the orders. The revenue was huge. Yeah, it was quite fun. It was quite a night. In the morning, we end up drinking a lot of Red Bull. And then everybody was sleeping all over the place. We see the CEO was sleeping over there. The CTO was sleeping over here.

And yeah, it was quite a night. I think it was 2017, I think. The 2018 has got a completely different story. 2018 was different because Alibaba got involved. Alibaba had bought Lazada in 2016, but it had mainly left it alone. Southeast Asian software companies were used to this, to being left to do things their own way. When you're working in Southeast Asia, you need to move fast. But at the same working in Southeast Asia, you need to move fast. But at the same time, in Southeast Asia, you have six different countries speaking six different languages with six different cultures.

How the hell do you move fast? That's very difficult. Instead of building one website, now you have to build six different websites with eight different languages. And at the same time, you have to deal with regulation of six different countries. For example, Indonesia,

they have laws where, similar to GDPR, where the data cannot leave the border of the country, the physical border of the country. So now you have to go into the country and build the data center. And this is back before AWS

was like a thing over there. So you have to go in there, you need to establish a data center, and So you have to go in there, you need to establish a data center, and then you need to deploy a different stack, an isolated stack compared to like the rest of the country. And then now you have to handle data synchronization between different tenants in your system, different data center. It makes the problem a lot more complex. And overcoming this complexity gets you ahead. It's reasons like that that Uber and many other hyper-growth companies

lost money going into Southeast Asia. And it's not just American companies who are out looking for growth. China has tech giants as well. And Alibaba is a big one. Alibaba is the owner of Timo and Taobao. That's the equivalent of Amazon. One is C2C and one is B2C. They own AliCloud. So if you want to do business in China, you need to host that somewhere. And that somewhere is AliCloud. Amazon footprint over there is very small. The comparison of Baidu and Tencent are

very small. The most reliable one is AliCloud. Most of the major banks are tenants of AliCloud. So most of the major banks out there are hosting their services on AliCloud. So they were very big. Aside from that, they do various different business. For example, Yoku is a video streaming platform. That's the size of YouTube. They also operate something that most people don't know about called China. And China is a logistic platform that handle all the shipping. So stuff that you are buying on AliExpress today, or even on Amazon today, most likely that was shipped over from China to your country, to your port, with China. That's a global logistic platform that they acquired. That's very big. So they were all over the place. They were really big. They are a big tech giant.

The acquisition caused some friction at Lazada. The company made the decision sometime around 2016, and that's very early, to move to use Go and Kubernetes instead of deploying on bare metal VM. So everything was very modernized. Our throughput was not that high compared to the like of Google or Amazon, but definitely it was one of the highest services at the time in Southeast Asia. There were really a lot of change and change come with

uncertainty. That's innovation. The rate of innovation was really high. So the platform in turn, you know, becomes a little bit less reliable. That's just a business risk that the entire group willing to accept. But once Alibaba came over, these folks are highly experienced. They are used to handling, you know, 100 times, 1,000 times higher traffic without sacrificing that resiliency. So they decided, hey, we want the system to be more reliable. Let's replatform everything to something that we know that's going to work. For some developers, switching tech stacks was a big ask. It's throwing away a lot of knowledge.

And Alibaba's stack is all Java and Spring. Everybody was like, Java sucks. We like Golang. Get the hell out of here. Where's your Kubernetes? Oh, you don't have Kubernetes. Where's your Docker? Oh, you don't have Kubernetes? Where's your Docker? Oh, you have your own Docker?

Nah. But it turns out, you know, there are good stuff in Chinese tech that's worth learning about. And I think some of that Westerners still don't know about today. And yeah, it was quite an eye-opener. An acquired company being forced to take their working software

and replatform it to however the parent company does things. It's a classic failure mode of acquisitions. The replatform will take years and you'll end up with worse solutions because of assumptions built deep into how the acquired company did things. But in a bigger way, replatforms just fail because they end up being bigger investments than the company thought they were going into it. And here's the thing about Alibaba. Alibaba was not afraid to invest in Lazada. That's just something that the company

decided and they set target and they actually put quite significant investment into it. Later on, I learned that this project is actually the biggest project in Alibaba at the time. What they called it at the time was a C-level project. And a C-level project meaning that the CEO of Alibaba was actually quite invested into it. And they managed to extract all of the talents from all different business units in Alibaba into a special team to make this happen. And at the time, I couldn't understand why the hell would you invest in Southeast Asia? But later on, when I started working for Alibaba, I started realizing, oh, this is actually a much, much bigger picture than I ever imagined. Oh my God. Like turns out like it related to a long term

planning that pivoting with the Chinese belt and road initiative. And yeah, it was way bigger than than we ever imagined. And it all makes sense after like we learned about that. But before that, everybody was like Java sucks. So everyone at Lazada starts learning about Alibaba Flavored Spring. And for San, this was a chance to learn about a whole different world of software development. How things worked at the scale of Alibaba Group. They flew a lot of us to China to work with the team over there to learn the text type in Alibaba, as well as to recode everything using the same business logic, but in Java, Spring Java of Alibaba. Alibaba does things their own way. They have their own JVM. They have forks of Git. They

have forks of MySQL. So there is a lot to learn. But Sun's team was an SRE team. They didn't own any code. So for 911, we merged with a team called GOC, Global Operations Center in Alibaba. And turns out, GOC is the overseeing eyes of the entire Ali group. It was essentially a thing of SRE team in Google. It's the centralized SRE team in Alibaba who, you know, measure all the monitoring system and handle all the escalation from different business unit and route it back to tech side. So now Sun and his team are SREs for Alibaba group. But there's a problem. The internal software used is all in Chinese, which Sun can neither read nor speak.

So his new project involves a lot of translation. So all the UI, all the system was in Chinese. All the documentation was in Chinese. We were over there trying to build new features. And yeah, it was quite fun. It was quite an experience. I get to spend one month in Hangzhou, in Alibaba campus. And boy, it was quite impressive. So is there a language barrier? Oh yeah, definitely. They put us in a hotel and we don't speak the language. So we couldn't order

foods. We were ordering ordering stuff by pointing fingers. Turns out you cannot get Google Translate in China, by the way. So putting up your phone is not even an option. Later on, I learned that Microsoft Translate actually works. So I need to download Microsoft Translate instead. But it's still quite difficult to communicate in China. And that's just living working is even more difficult everything need to go through translation all the documentation

and sometimes the translation would be wrong and then over time you need to learn the dialect of the translation like for example they'd like to translate a use case into something called scenes, like scenes in the movies. Every time you want to talk about use cases, they would translate it into scene. I was like, what the hell is a scenes? And turns out later on, I was like, oh, they meant scenario. So over time, you just like learn, pick up like little, little things and everything start making a lot of more sense. But yeah, everything was like very difficult. Yeah.

But like, how did it work to code? If you were working on this ticketing system, but the other guys didn't speak, you know, English or Vietnamese? Some of them do speak English, but not at the level where you can like communicate technical with them. We have to let the code speak for itself. Now, fortunately enough, I was blessed with a very talented teammate at the time. And he was able to navigate through the entire full stack ecosystem,

like from React.js down to Spring Java. And he was impressively retouching me a lot of that. But the app itself was quite simple. Think about a ticketing system. It's just CRUD. And then you can attach like several metadata. You have a state machine to handle different statuses. The most difficult part there is integration.

Alibaba is a huge ecosystem. You have a lot of things to integrate. For example, you don't have Slack over there. You have DingTalk. DingTalk is a chat system of Alibaba. They also have what they call a change management system so that all the changes in Alibaba, all of the deployment, all of the network changes can be automatically revert by one click. So the moment when you deploy and something goes wrong, that deviate from the metric baseline that the AI engine put out, then you have an option to click one button and revert all that change immediately. From database schema to deployment to migration, everything, network changes, one click of a button.

But yeah, it was like a lot of fun. The team Sun merged with was also ops for Alipay, which became its own problem. So in China, if you live in a megacity, people don't accept cash. Like when we go over there, we have a really hard time paying for food because people would only accept either we pay or alipay we couldn't go out and buy bubble tea because

people were like oh that cash i don't want to bother having to go to the bank and and you know put this into the machine like that is just too much work. Alipay dude. And we were like having to beg our teammates so that they would like exchange cash that we got from Vietnam. And these are Chinese cash. It's not Vietnamese cash. We were like begging them, hey, could you take this, take this to the bank and then just transfer some of your Alipay credit? So yeah, using Alipay was a challenge, but a bigger challenge was just the scale that the Alibaba group operated at.

Three rooms away from my rooms, it's a network operation center. So that's where you have like the big TVs with all the graphics. So our Chinese teammate was sitting in there watching all the alerts, watching all the graphs. And yeah, like it was very cool. You can see the auto-throughput of all the platforms, all the traffic, all the networks.

And boy, is it massive. We were very proud of the traffic that we have in Southeast Asia. It's not even a drop in the bucket compared to China. Turns out when you have billions of people speaking the same language, watching the same live stream, having the same culture, turns out that creates like a huge traffic spike. Imagine you put a laptop on sales for $100 and you have 10 stocks. Often in time when you build an e-commerce system, that would be like one row in your SKU database, the stock keeping unit. And then that one row, when somebody ordered

it, it will change the status to sold. What happened if 10 billion people were trying to place an order on a single row in database? Like, how do you design a system for that? It's an interesting problem, right? How much is reliability worth at scale? And how much does it cost? How do you build for 11.11 day when you have a billion users all at once? Things was massively impressive in China.

The way they have to re-architect things in order to accommodate for the load that they were having. I just don't see the same thing in Europe. I don't see the same thing in North America. Even like the population in North America, like US and Canada combined, it's not even half the population of China. So the load we are talking about here is massive. The scale is completely massive and they don't have access to Google. They don't read Stack Overflow. So they have their own stuff.

When I was over there, I had to learn all of my own stuff over there. And I was like, Oh my God. They just went completely different route. We were like very fascinated with Golan, Kubernetes. These guys were like Java, Java, Java, Java, Spring Java, Spring Java. They just like completely innovating without us knowing in Western world. Nobody was discussing how China was building stuff. Turns out they forked their own MySQL.

They were forking their own JVM. They forked their own Nginx. They forked Git so that it would use their internal object storage platform instead of on disk. So Git hosting everywhere in GitHub GitLab required you to have a disk, a physical hard drive. But these guys were forking Git, inserting libgit2 into it, and then let libgit2 talk to their internal implementation of S3. And yeah, they were able to scale their DevOps platform to a cloud-native scale before anybody else can. Even today, like GitHub and GitLab still struggling decoupling storage and compute.

And these guys already have that. Similarly, they do the same thing with MySQL. They built on top of the knowledge they learned from Facebook about forking the storage engine underneath MySQL, they built on top of the knowledge they learned from Facebook about forking the storage engine underneath MySQL to RUBDB. And then these guys were running into a problem where RUBDB is an LSM3 database storage engine. And the problem with that is that the compaction of these different layer of logs was getting quite slow and consume a lot of CPU. So guess what they did?

They offload that computation to an FPGA. Like these guys were like running custom hardware, custom chips, so that their database can handle the load better. And that's just massively impressive. That's just like not something that you can find in any tech company. Custom chips. What the hell? Another surprising thing at Alibaba was just how much everybody worked. 996.

When I went over there, everybody told me about 996. You worked from 9 a.m. to 9 p.m. six days a week. So Monday to Saturday. And that's pretty much the life over there inside the office is very common for you to see beds even camping tents yeah no seriously like in alibaba you on in the hallway you would be able to find several camping tents gender separated so that folks can go to sleep. Not at night, but like middle of the day after lunch. Sleeping in Asia is quite common.

But especially during campaign night, like Double 11th, people stay over at the campus all the time. 996 works out to 72 hours per week. But don't worry, nap time's included. At first, Sun didn't like or understand this at all. But later on, I started realizing why it is the way it is. If you're a tech company and you tell people to work nine to nine, six days a week, that's not sustainable.

Nobody's going to work for you. So be able to tell people to work from nine to nine, six days a week, and people still wanting to line up to work for nine to nine, six days a week, and people still wanting to line up to work for you, there must be something special there. And turns out later on, when I start talking to my Chinese teammates over there, they start telling me stories, stories that make quite a lot of sense. So think about it like this. What do you do after you gain a stable job? You get married. So before you get married in China, my friend was telling me this,

you need to have at least a house and a car. And if you go work for Alibaba, because they also co-own a lot of Alipay, which is a financial company that people were using instead of bank and cash, they also do loans. They can loan to you your mortgage at a very low interest rate. So the more senior you are in Alibaba, the better interest rate you get. Similar to car, you can loan from the company to buy a car. And that, the more senior you are, the better interest rate you get. And that's not the only thing. Turns out if you work for a big company, your credit score is very high.

So credit card provider would actually try to line up outside of the company so that when you left, they can hand you flyers. Here are free money. Please loan from us. So that's the first thing. The second thing is the lifestyle on campus. Everybody was eating in the cafeteria. They have nine different buildings and four different cafeteria in the nine buildings.

So the food was being mass produced. So they were able to produce that very cheap, right? So they were able to sell, provide this food to the employees very cheap. Now, I haven't worked at a big tech campus, but some of this is starting to sound not that dissimilar from like a meta campus or Google campus. Sun took a lot of photos while he was there. And if you want to see some of them, I've shared them on Twitter. Tents to sleep in, massage stations, places to eat, and just the scale of things. But yeah, people weren't heads down working all the time, and the company had lots of perks to offer them.

So workday ended at 6 officially, but the food between 6 and 7 p.m. would be sort of a discount. So employees are incentivized to stay on campus after 6. And then at 9, they could sell you free snack. So single employees who have nothing to do at home can stay on campus until nine to get those free snack. And taxi, because Alibaba was owning that taxi company, if you go home, now you get free taxi. Oh my God. Now you start seeing all the integration start coming into play. Hey, these guys actually provide a lot of incentive for people to work 9 to 9.

And then the next thing you know, if you need to go out and buy anything, hey, Timo and Taobao will deliver that to the doorstep of your building. You get faster shipment on campus compared to going home. So you would route all of your e-commerce purchases to the company address. Just tell it like what building is it and then go down and get it. So it's more convenient to live on campus compared to going home. So people was like getting all the perks everywhere that affecting their life. They willing to work for it. All these perks means a lot of people want to work there.

And then the higher you climb the corporate ladder, the more perks they are, which leads to the fact that internal competition can get pretty intense. They're willing to compete for the jobs. And that's why Alibaba had a very cutthroat culture so that they can digest out low performers every year i think like 30 of the low performers digest out and everybody was like competing for the top because turns out

one more perks which is your final bonus your yearly annual bonus can be up to 100 of your salary oh wow that's huge so people was competing to be the top performance. So these guys know what they were doing. How do you manage a giant company of software developers who all want to work really hard and all really want to advance their careers? Well, maybe you already know the answer. The way Alibaba managed is very top-down, very KPI heavy. So every business unit are given a certain KPI. The leader of those business units would assign KPI to each of the sub-departments. And each of the sub-departments would assign KPI down to their employees.

So they have metrics to meet. These guys were very metric-driven. This KPI or OKR system, much like any top-down system thrown at motivated people, leads to some unintended consequences. For instance, porting Lazato to the kernel platform was one thing, but actually running the company? Maybe that was a less exciting project. And this is, you know, like this hearsay.

So I'm not saying this as a source of truth, but they were rumors to be like, they think Lazada to be a not profitable place for them to meet the KPI. So nobody wanted to be a CEO of a failing position. Lazada actually changed CEO quite a lot because nobody was meeting their KPI. So the culture there is quite cutthroat. And people were competing for nice position so that they can be set up for success. There's a way it sounds very foreign. There's another way that it sounds like Google.

Well, I cannot tell you because I have never worked for Google. But what I can tell you is that they move very fast. And I don't think anybody has been moving as fast, especially if you work in a Western country, in European or in North America, you're not moving fast. These guys move a lot faster. Okay, so Alibaba in China moved fast. And that can be exciting But for Sun, it had downsides

Alibaba obviously providing a lot But I was interested in learning what else is there And I was a little bit burned out With all of the Google Translate of Chinese documentation Yeah, I was looking for other opportunities And my boss was like Hey, I'm going to the exact same company, but in Singapore. So Sun stayed working for Lazada, but moved to Singapore,

where his time on site made him the expert for his team. And then I get to teach my teammates about Alibaba tech. They were like using Python at the time. A lot of Python, Django, Flask, self-hosted on AliCloud. And I was like, no guys, here's AliJava. Here's all the goodness. Hey, imagine Spring Java, but everything come with battery equip. You want a database?

Here's a database. You want a message queue? Here's a message queue. You don't need to worry about infrastructure. So everything came out of the box. And that was actually a downside because I started to realize that the world outside was moving quite fast. And even though I was quite an expert in internal technology in Alibaba, I know nothing about the world outside. Sun wanted to keep growing and learning and not

pigeonhole himself into one company, into one tech stack. So he set a deadline for himself. So I set out a plan that after I arrive in Singapore for a year, I would need to learn new technology and explore the job market. At the time, I was actually interviewing with Google in Singapore. I attended several of their tech talk, get to talk to the GPay team over there in Singapore, and they were doing a lot of interesting things. Google famously has a super long hiring process. And in the meantime, Sun found an opportunity to do something completely different. At the time, I was like, you know, getting to know my girlfriend at the time and now my wife. And I was like, OK, I need to settle down someplace.

And good luck buying a house in Singapore. Singapore was way too expensive. Your best luck is being able to rent a nice apartment. And that's pretty much it. You would not even get a car. It would cost you millions to get a car. And that's pretty much it. You would not even get a car. It would cost you millions to get a car. And that's insane. Sung got an offer from Booking.com. Booking.com is an online travel agency based in the Netherlands, in Amsterdam. And in some ways, Amsterdam is the

opposite of Singapore. I took that opportunity and said goodbye to Singapore, even though I love Singapore so much. It's such a nice country. Wonderful. Strongly recommend. Where's the food better? Oh, Singapore. Definitely. Netherlands is like the worst. Oh my God. No, seriously. Like even, even Dutch hate the food in the Netherlands. So I left Singapore for the Netherlands, August 2019. And then me and my wife officially get married like on the cubs of COVID. I remember my product manager on the first day I came to work, he sit me down with my team lead. My team lead was actually joining with me.

We did the orientation together and my product manager sit both of us down and explain. Hey guys, booking paid a lot of money to relocate you guys here we want you to success here that means you need to prioritize your life first you need to be able to find a house to rent you need to be able to do all your paperwork so in the next upcoming month if you need to take time out, just go and then come back and tell us later. And that is such a huge opener because that's just a mental shift, completely 180 compared

to Alibaba because Alibaba, everything was work, work, work. Your life is in the company. Your metric is important. Booking was not like that. Booking was like, no we we respect you as a human we know that you have a family we know that you have needs take care of them and then come back to us work your best right and that's just changed everything in booking i have teammates who are blind i have teammates who are gay lesbian disabled We were all able to hang out together.

No problem. There's no discrimination. It's just a huge open and welcome culture. Very diverse pool of folks over there. And yeah, I just love it so much. So we still have, you know, massive traffic. We still have problem at scale that we have to take care of. And yeah, it was very fun trying to find work-life balance

in midst of all that, in midst of COVID, in midst of relocation, and in midst of new marriage as well. So for now, Son has found his place in the world. He and his wife are still in the Netherlands. He found his home and a place where the work culture works best for who he is. But he only knows that because of the experiences he had. If he had started at booking first thing, maybe he wouldn't have realized what he had because he'd have no comparison. And that fact shaped the advice that he offers to others. Life is diverse. Before I get to work in in china it was a black box to me it was not occurred to me that there was a payment system that's so advanced

that people would refuse using cash and yeah i know credit card exists but that doesn't mean like when i hand you cash you was refused to receive it. Just at the same time, going to Southeast Asia, seeing various different startups trying to get themselves a place on the map was hugely game-changing because it enabled me to think in terms of business, in terms of thinking, how would I operate this if I were the CEO or CTO? I also learned how to party. In Waterloo, they drink bubble tea. You can get way better party going to Toronto than trying to do it in Waterloo.

Going back to Vietnam, every one week or two weeks is a drinking party. A lot of beers were consumed, a lot of hot pot. The food is wonderful over there when i go to singapore i learn what good planning can bring you can just enter the city and you can see like the first thing you see in singapore is the best most beautiful looking airport in the world and you can see like oh my god how much would it take for my government to build this airport. Infinity money would not change this because this is not a problem with money. There's a problem with planning.

And then Netherlands, they taught me to be open-minded. They taught me about work-life balance. And all of that is just come back to be travel more. This is not an advertisement for Booking.com, but travel more and learn more from it and and keep an open mind because yeah like maybe western media can can tell you a lot of shit about china about how evil things can be but try to go over there and try to see how people leave being able to see like thousands of people eating lunch all at once, surrounding you, all looking like you, because I'm Asian, make you humble. It does have a humble feeling entering that cafeteria and see,

you know, hundreds of people lining up for the same dish. It makes you feel insignificant. So it's a massive war out there. Travel more. That was the show. Thank you so much to Sun. You can find him on Twitter. I'll put a link on the episode page. And in an upcoming bonus episode for podcast supporters, Sun is going to share his views on the future of Chinese software development and a little bit about the Chinese expansion via the Belt and Road Initiative. It's super interesting stuff.

A lot of the innovations that Alibaba is doing are out there on GitHub and they're open source and they're for the taking. But outside of like certain system researcher circles, it feels like nobody's paying attention. That's gonna change. So support the podcast and on Patreon, you can get access to that episode that I'll release in the future.

Also, Sun's career keeps evolving. So you'll hear a little bit about his latest career pivot, which is how I ended up meeting him. Also, if you'd like to hear a little bit of the fun backstory behind the episodes, then this will be a great opportunity to hear that. And until next time, thank you so much for listening. Thank you.