Dingsheng's Outsourcing

The next morning, the first thing Han Luyi did after waking up was turn on his computer and reopen the outsourcing request.

Last night, he had not made a decision. After sleeping on it, he had already thought it through.

The technical summary attached to the request page was only three pages long. It had been sanitized, with the company and project names removed, but to Han Luyi, it might as well not have been.

He could match every description to a specific code file.

"The core recommendation engine has recently undergone architectural adjustments, replacing the original dynamic weight allocation module with a rule-based static engine."

Han Luyi's expression turned subtle when he saw that sentence.

Dynamic weight allocation was the core scheduling module he had spent eight months designing—the heart of the entire recommendation system. Its brilliance lay in its ability to automatically adjust recommendation weights according to users' real-time behavior. Every click, every swipe, and every pause was factored into the calculations.

That system was not something that worked the moment it was written. He had coded and refined it at the same time, running hundreds of millions of user data records and grinding through the parameters one by one.

Then Chen Bowen's team had torn it apart and replaced it with a static rule engine.

Han Luyi understood why. The dynamic weight module was indeed complicated. Without him there, no one else on the team could fully understand that logic. Chen Bowen's choice had been simple: if he could not understand something, replace it with something he could.

The technical summary also listed a series of symptoms: response latency surging during peak periods, declining recommendation accuracy, continuously falling user retention, and A/B tests performing worse than the old version across the board.

To put it bluntly, everything had gone to hell after the replacement—and it was only getting worse.

Han Luyi leaned back in his chair and activated Sight.

His energy was full: 250/250. He focused his attention on the three pages of technical summary on the screen.

The panel exploded open.

This was no ordinary popup. The entire screen was covered by a translucent red warning, like a wall on fire.

[Target System: Dingsheng Digital Technology · Core Recommendation Engine]

[Overall Stability Rating: Extremely Dangerous]

[Estimated Time Until Total Collapse: Within 72 hours of launch]

One hundred and forty Bugs, some deep red and some pale red, densely filled the screen like a critical condition notice.

He only glanced at the first three critical Bugs before he no longer wanted to look. A memory leak—the system would choke itself to death in three days. A broken data pipeline—one-quarter of user data was quietly being lost. A concurrency deadlock—a twenty-percent chance of an outright collapse during peak traffic.

Three ticking bombs. Any one of them exploding would be enough to make the trending topics.

He had spent three years refining this recommendation system from merely usable to genuinely good. Every parameter, every weight, every version iteration had been typed out by him, line by line.

Chen Bowen had only taken over half a month ago.

He remembered the line on Chen Bowen's panel on the first day he had been laid off—

[WARNING: This target has 3 critical Bugs. Click to expand. (Current level insufficient)]

Back then, he had been Lv. 1 and could not see the details.

One of those three critical Bugs was probably this. Chen Bowen had picked his fruit, but he could not chew it.

Of course Han Luyi could fix it. He had written this code himself. He did not even need Sight; for some Bugs, one look was enough for him to know which line needed changing.

He deactivated Sight.

The screen returned to the normal outsourcing request page. In the upper-right corner were two buttons: [Accept Request] and [Ignore].

Han Luyi's finger paused over the mouse.

Accept it? Save Chen Bowen? Am I an idiot?

Then he clicked [Ignore].

He would not take it.

He closed the outsourcing page and took a sip of the coffee on his desk, which had already gone cold.

Not taking the job was one thing, but it would be a waste to keep some things bottled up.

He opened a forum and, as NullPointer, spent forty minutes writing a purely technical post titled <Seven Classic Pitfalls in Recommendation System Architecture Refactoring>.

Every one of those pitfalls was something Chen Bowen was currently stepping into. But Han Luyi knew that, at Chen Bowen's level, he would not understand it even if he saw it. When the system truly crashed and this post was dug up again, it was hard to imagine how spectacular Chen Bowen's expression would be.

After posting it, he closed the forum.

His mind began turning.

These 140 Bugs, the more than one hundred Bugs he had taken jobs to fix over the past two weeks, and before that, the thousands of Bugs he had handled over five years.

They looked different, but their patterns were all the same.

The same patterns kept appearing in different projects. Every time he fixed a Bug, he was essentially doing two things: recognizing the pattern and matching it to a solution.

Wasn't that exactly the kind of work AI could be fed?

He put down his cup, took out his phone, and opened the note containing a single line—

"Code review tool, connecting code with real-world scenarios, BugKiller?"

He could organize every Bug pattern Sight had identified—the features, trigger conditions, and repair paths—and feed them to AI. Teach it to "see the code and know what problems would arise after launch." AI did not need Sight; it only needed to learn from what Sight had seen.

It was like teaching AI to play chess. You did not need to teach it how to think; you only needed to show it enough games.

Han Luyi turned on his computer, created a new folder named "BugKiller," and created a new document: Bug Pattern Database v0.1.

Then he began to write. He wrote quickly. Every pitfall he had encountered over five years poured out from memory and automatically fell into place by category: memory, concurrency, data flow, interfaces. Under each category were trigger conditions, typical symptoms, and repair paths.

It was as though he were compiling a Complete Guide to Bugs for a novice who knew absolutely nothing.

After writing for two hours, he stopped.

It was not enough. One person's experience could not support an entire database. He needed more samples and more systematic classifications.

He needed a product manager.

Han Luyi picked up his phone and found Su Niannian's WeChat chat window.

"Nian Nian, I have an idea about the AI code review direction we discussed last time. Are you free tonight? Let's have dinner and talk."

Less than thirty seconds after he sent it, Su Niannian replied, "Tonight?! What time?"

"Seven."

"No problem! I get off work at six-thirty anyway. What's the idea? Tell me first?"

"I'll tell you when we get there."

The "The other party is typing..." indicator at the top of the chat window flashed several times. In the end, what she sent was not another question, but:

"Then I'll pick a restaurant first. What do you want to eat?"

"Anything. I'm not picky."

"Okay, I'll decide."

Han Luyi found it a little funny. He had been the one to invite her, but after three sentences, she had taken control.

But that was normal. Su Niannian was a product manager. Seizing control of requirements was instinctive for her.

Han Luyi put down his phone and buried himself in the document again. By the time he looked up, it was already four-thirty in the afternoon.

There were still two and a half hours until seven.

He looked back at the BugKiller folder he had just created on the screen. After thinking for a moment, he decided not to continue writing. When he spoke with Su Niannian tonight, he needed living ideas in his mind, not words stiffened on a page.

He shut down his computer, got up to take a shower, and changed into a clean shirt.

As he left, Lin Duoduo's voice came from next door as she told Lin Wanqing about kindergarten.

Han Luyi closed the door behind him and headed toward the subway station.

The evening breeze was neither cool nor warm. In mid-March, Haicheng sat right in the crack between two seasons.