Chapter 9 Interview
Chapter 9 Interview
In the morning at Kexing Science Park, sunlight streams through the glass curtain wall into the third-floor corridor.
Lin Shen arrived forty minutes early. After registering at the front desk, he received a temporary visitor card and was led into a conference room.
There were already seven or eight young people sitting inside. Some were muttering something to themselves, while others were repeatedly flipping through their resumes. A faint sense of tension permeated the air.
He found a seat by the window, but instead of looking at his resume or reciting key points, he stared at a moth repeatedly hitting the glass outside the window.
"Why do you think it insists on crashing into that spot?" he suddenly said to a boy next to him who was memorizing algorithms.
The boy paused for a moment, then looked up and asked, "Huh?"
"There's an open window right next to it," Lin Shen pointed to the end of the corridor, "but it insists on crashing into this spotlessly clean glass. Isn't that like some people who, even when there's a path, choose the most difficult one?"
The boy opened his mouth, completely at a loss for words. He even wondered if the person was mocking him.
Just as she was preparing a barrage of verbal rebuttals, she saw Lin Shen had already stood up.
Lin Shen walked slowly along the corridor. In 2010, Tencent's Kexing Campus didn't yet have the cafes, gyms, and sleep pods that would later appear; everything seemed simple and pragmatic. The slogan on the wall remained unchanged: "Everything is based on user value."
The break room is probably still the same, isn't it...?
He walked to the end of the corridor and, sure enough, saw a small tea room.
There was no one there, only the low hum of the coffee machine and water dispenser. Lin Shen easily found the stock of solid juice sticks, made himself a cup. He never drank coffee when working overtime late into the night; when he felt tired, he would make some juice to perk himself up.
As I was leaning against the bar counter, slowly drinking, a man and a woman walked in.
They were all wearing Tencent employee badges. They didn't look at Lin Shen, but just poured themselves a glass of water and chatted. The man's tone was obviously irritated.
"...Since I was temporarily chosen to be an interviewer, who's monitoring the progress of the projects I'm currently working on?"
"Teacher Zhou said that these interns will be given special attention and will be assigned to front-line teams in the future."
"Training? What can you learn from just a few written test questions? Last year, the Peking University student who got a perfect score on the written test couldn't even detect a simple memory leak."
Lin Shen's fingers tightened slightly as he gripped the paper cup.
He recognized the voice... it was Chen Mo.
In his previous life, he had only met Chen Mo a few times at internal technical sharing sessions, when Chen Mo was already the senior technical director of the WeChat business group. Now, Chen Mo, in his early thirties, was tall and straight, with slightly disheveled hair, and spoke with the obvious straightforwardness and frustration of a front-line engineer.
"You're being a bit extreme. I've seen that person doing quite well in the entertainment division. Speaking of the written test, Professor Chen, the design question you set was way too difficult. Perhaps you should have chosen a different one."
"I insisted on adding the communication question," Chen Mo said to his female colleague beside him, his fingers tapping unconsciously on the table. "What the team needs most right now is someone who understands low-level communication. Those who can only write UI will have to be taught from scratch when we hire them. I don't have time to train interns like that."
The female colleague shook her head and added, "A 40-point question, the average score is expected to be less than 15. Do you expect a recent graduate to design a usable IM architecture?"
"Difficulty is what differentiates people." Chen Mo pushed up his glasses. "I don't need people who can just memorize answers; I need people who can truly think critically."
Lin Shen subtly put down the paper cup and prepared to leave.
Eavesdropping on the interviewers' discussion is inappropriate.
But just as he passed by the two of them, Chen Mo suddenly turned around: "You...are here for an interview?"
The sound wasn't loud, but it was very clear.
Lin Shen stopped in his tracks and nodded frankly, "Yes. Sorry, I didn't mean to eavesdrop on your conversation, I was just about to leave." After he finished speaking, he added, "But regarding the memory leak you just mentioned, sometimes it's not that we don't know how to check, it's that we're afraid to check. What if it's a leak in our own code? That would be so embarrassing."
Chen Mo and his female colleague were both stunned.
That's too blunt, even a bit...provocative?
But Lin Shen wore an innocent expression on his face, as if to say, "I'm just stating the facts."
"What time is the show?" Chen Mo asked, his gaze lingering on Lin Shen's face for a second longer.
"Ten o'clock."
"The first one?" Chen Mo turned his gaze to his female colleague beside him.
Lin Shen later learned that her female colleague's name was Li Wei. She opened the folder in her hand. Inside were the interviewees' information, arranged in chronological order.
Her gaze lingered on the first page for two seconds, then she looked up, glanced at Lin Shen, and handed the folder to Chen Mo.
Chen Mo took it and his gaze fell on the paper.
A few seconds of silence.
Then he looked up, his gaze towards Lin Shen holding something different: "You're Lin Shen?"
Lin Shen remained calm: "It's me."
Chen Mo closed the folder, but didn't let him leave immediately. Instead, he leaned against the bar counter and casually asked, "You got 192 on the written test and 38 on the system design. For that IM architecture question, you wrote a solution of 'weak push + timed fetch + smart wake-up'—why did you choose this approach?"
"Teacher Chen, the interview hasn't started yet!" Li Wei quickly said.
"It's alright, let him talk, there's still time." Chen Mo waved his hand, but his eyes were fixed on Lin Shen.
The problem came suddenly, but Lin Shen was prepared.
"Because of the real-world constraints of mobile devices," he said calmly, "limited battery capacity, unstable network, and strict system process management. Pursuing absolute real-time performance would lead to excessive power consumption and a degraded user experience. So I think it's better to accept a certain delay and use resources to ensure that important messages are delivered."
As Chen Mo listened, he tapped the table with his fingers again, as was his habit: "But users might want immediacy."
"So product design is needed to complement it," Lin Shen naturally chimed in. "For example, prioritizing messages, giving users choices, and explaining the purpose of background processes. If the technology can't be perfect, then the product and user experience can make up for it." He suddenly laughed, "It's like dating. You can't watch your partner 24/7, but you can agree to 'call immediately for important things and text for minor things,' it's the same principle."
Li Wei raised an eyebrow.
Chen Mo stared at Lin Shen for two seconds, then—smiled: "The metaphor is a bit strange, but I understand the meaning."
"The balance between technology and product..." Chen Mo repeated thoughtfully, then suddenly changed direction, "For the multiple-choice question about communication protocols, you chose 'layered design, choosing TCP or UDP based on the scenario,' and the examiner gave you full marks. But if you were to design it specifically, how would you layer it?"
Lin Shen thought for a moment: "Three layers. The first layer, critical state synchronization and important messages, must use TCP to ensure reliable delivery. The second layer, real-time audio and video streams, use UDP with forward error correction, tolerating some packet loss. The third layer, non-critical state updates, such as 'typing,' can go through UDP; if they get lost, so be it."
He paused, then added, "Actually, it can be refined further—for example, dynamically adjusting the strategy based on the current network quality. We can be more aggressive under Wi-Fi, and more conservative under weak networks. But this involves synchronizing the state of the client and the service." His eyes suddenly lit up. "That's right, it's like walking. You can run on a flat road, but you have to walk slowly on a gravel road, and on a rainy day, you have to look carefully every step of the way."
The tea room was quiet for a few seconds.
Li Wei glanced at Chen Mo, while Chen Mo stared at Lin Shen, his scrutinizing gaze gradually replaced by a certain interest.
"Forward error correction, dynamic strategies..." he repeated the two words in a low voice, "Where did you learn these concepts?"
"I've studied some open-source projects and read a few papers," Lin Shen said honestly. "But it's all still just theory; I haven't actually done it."
That's the truth. He had done it in his past life, but Lin Shen shouldn't have any relevant experience at this point in time.
Chen Mo nodded and didn't ask any further questions. He glanced at his watch: "It's 9:40. You can go back to the waiting room now; it starts promptly at 10:00."
"OK."
Lin Shen nodded politely and turned to leave the tea room.
On his way back to the waiting area, he could feel two pairs of eyes following him. Not scrutiny, but more like... confirming something.
Back in my seat, the sunlight outside the window had risen a little higher, casting brighter patches of light on the corridor floor.
The other interviewees in the waiting area were still preparing nervously; some were quietly reciting algorithm complexity, while others were drawing system architecture diagrams on paper. Lin Shen leaned back in his chair, closed his eyes, making the boy from before feel somewhat uncomfortable.
Lin Shen, however, was reviewing his brief exchange with Chen Mo. Although it only lasted a few minutes, it contained a great deal of information.
First, Chen Mo remembered his written test answers and was deeply impressed by them.
Secondly, Chen Mo was interested in the idea of "balancing technology and products"—this is very "WeChat" and very "Zhang Xiaolong".
Third, those advanced concepts, such as forward error correction and dynamic strategies, attracted attention but not skepticism.
well.
At 9:50, the receptionist peeked in and said, "Lin Shen? Meeting room 302, you can go in now."
He stood up and straightened his shirt collar.
On the way to the conference room, the workstations on both sides of the corridor were already full of people. The screens were flashing code, the keyboards were clicking rapidly and rhythmically, and the whiteboards were covered with complex architecture diagrams. This was Tencent in 2010, in its preparations for the eve of the mobile internet wave.
In conference room 302, Zhou Botao, Chen Mo, and Li Wei were already seated on one side of the long table.
Lin Shen pushed open the door.
"Lin Shen, please have a seat," Zhou Botao said, his voice as steady as during the phone interview.
"My name is Zhou Botao, and I am the chief interviewer today. This is Chen Mo, a key technical member of our team; and this is Li Wei, an HRBP. Our interview today will last about an hour and will cover several aspects, including technology, projects, and product thinking."
"Okay, hello teachers."
Zhou Botao opened the folder in front of him, his gaze falling on Lin Shen's resume, and began the interview: "Lin Shen, we have reviewed your written test, especially the system design question. Your thinking was very clear, and your basic skills were solid."
However, there's a difference between theoretical discussions and actual development. Let's start with the "simple weather app based on the Android platform" on your resume, and talk about the biggest challenges you encountered in actual development and how you solved them.
A standard opening for a technical interview is to assess the authenticity of project experience and problem-solving skills.
Lin Shen was prepared. Instead of choosing the most complex technical challenge, he opted for a problem that was more "story-driven" and involved a "learning process".
"The biggest challenge is not in the technical implementation, but in gaining a deep initial understanding of the characteristics of mobile development."
Lin Shen paused, organizing his thoughts, "At the time, following the habits of desktop development, I directly made network requests to obtain weather data in the UI thread. It ran smoothly on the simulator and my test machine, but when I sent the APK to another colleague who was using a low-end device for testing, the application frequently displayed 'Application Not Responding' pop-ups, and even crashed outright."
Chen Mo raised an eyebrow slightly, seemingly surprised by this angle, but also intrigued.
"Later, after investigation," Lin Shen continued, "we discovered that it was caused by a combination of network latency and insufficient UI rendering capabilities on low-end devices. I realized that the mobile development environment is 'harsh' and diverse, and we cannot use the devices that developers have as a standard."
The solution was to introduce asynchronous tasks to handle network requests and provide an explicit loading status before data is returned. They even considered dynamically simplifying some UI animations based on device performance. This experience truly made me realize that "performance" and "experience" on mobile devices aren't bonuses, but rather the minimum requirement. He suddenly tilted his head. "It's like hosting a dinner party; you can't just order according to your own taste. You have to ask the guests if they can eat spicy food, if they have any dietary restrictions, even if you might think, 'How could anyone not eat something so delicious?'"
Zhou Botao's lips seemed to twitch almost imperceptibly.
"Yes, learning from mistakes is the right direction," Zhou Botao nodded, then followed up, "If you were to refactor this application now, what would you optimize first? And why?"
"Data caching and update strategy," Lin Shen said without hesitation.
"Weather data doesn't need to be absolutely real-time, and frequent requests consume both power and data. I will design a local caching mechanism that intelligently decides whether to initiate a new request based on the user's location and the last update time, and pre-cache potentially relevant city data in a Wi-Fi environment."
This involves a comprehensive trade-off between business logic (timeliness of weather data), user experience (refresh wait time), and technical implementation (caching effectiveness, storage space). He thought for a moment and added, "It's like your mom always trying to feed you, but you're not actually hungry—there needs to be a mechanism to tell her, 'Mom, I'm really full, I'll eat later.'"
Even Li Wei couldn't help but glance at him this time.
The answer was well-organized and naturally elevated the technical points to a level of thinking that combines product and technology; however, the metaphors... were a bit too colloquial. Zhou Botao made a few notes in his notebook.
At this moment, Chen Mo took over the conversation, his tone more formal and in-depth than when they were in the break room: "You just mentioned 'dynamically adjusting strategies based on network quality,' which is also reflected in your written test design."
Let me present you with a real-world scenario: Our service detects a sudden surge in message delivery delays for a large number of users in a certain area, but the local network operator reports that the network is functioning normally. As the engineer responsible for the communications module, what would your troubleshooting approach be?
novellhall