Skip to main content
  • Home
  • Library
  • Read
    • Popular
    • Charts
    • Browse
    • Categories
    • Wikis
    • Glossary
  • Request
  • Activity
  • Leaderboard
  • Help
Our Apps
  • DaoSearch
  • Narratorr
Social
  • Discord
  • Patreon
The AI Wave Tycoon's Rise
Chapter 102

Turning the tables

2,681 words 9 min read 2 views 1mo ago Source ​
Skip to the end of the chapter
Prev
Next

The changes at the Qiantan office began with coffee beans.

In Shen Congyun's administrative budget application, the amount of coffee beans increased from five kilograms to fifteen kilograms per month, followed by a series of new items: fifty mugs, three boxes of instant coffee for backup, and a certain amount of snacks. When Han Luyi signed, he glanced at the office area and saw that more than half of the workstations were occupied by unfamiliar faces, even though he had conducted final interviews with each of them and seen their profile cards.

New employees arrive almost every day. The front desk's access control system and Lark's onboarding process are running simultaneously, and Shen Congyun's assistant can't handle it all by herself, so she temporarily pulled in an administrative intern to help.

Li Ting was promoted to Operations Team, and she was in charge of two new employees, one in community operations and the other in user growth. It was the first time she assigned tasks to someone, and she thought for a long time about how to phrase it, but in the end she realized that she could just say it directly. The new employees were even more nervous than she was. The three of them started Starting Point the very next day.

Chen Jianye was promoted to chief architect, a title that sounds quite impressive. He had mentioned to Han Luyi and Shen Congyun that he wanted to try a management position, since at Dingsheng, only by becoming a manager could one truly advance. However, Shen Congyun and Han Luyi had discussed it privately and agreed that he wasn't suited to leading a team yet, and should focus on developing his architectural skills first. On the day of his promotion, Old Chen treated a few former colleagues to milk tea, his expression a mix of joy and disappointment.

Jiang Yixin officially transitioned to product manager and reported to Su Niannian. On the day her Lark signature changed from "development" to "product," she took a screenshot and sent it to her best friend, Big Liu.

[Big Liu]: Our Jiang, from now on we are class enemies!

[Big Liu]: Before that, let's have dinner first!

Jiang Yixin saved the screenshot to her favorites.

Product and development have also been further divided. Product team: Su Niannian, Jiang Yixin, and Add One newcomer; Open Things technology team:

The BugKiller technical team was led by Chen Jianye; the Heavenly Craft model team was Zhao Wenyuan's territory, with Luckin delivery bags always piled up at the entrance; Operations Team was under Li Ting.

Five groups, like five fingers, and the company finally looked like a company.

Han Luyi was the last to enter the small conference room. He looked around and saw that everyone else had already arrived: Lu Mingzhou, Su Niannian, and Jiang Yixin.

Su Niannian turned on the screen mirroring. The first image was the Daily Active Users curve for the past week, a zigzag line that shot straight from the bottom left to the top right, so steep it didn't look real.

"Daily Active Users exceeded 100,000, and more than doubled in a week."

The collapse of All Things Alive, Ma Xiaofei's videos, and CodeMonkey Old Zhang's Zhihu articles—three waves of traffic generation overlapped. New registrations returned to over two thousand per day, with about 40% traceable to social media sharing links. The paid conversion rate increased rather than decreased because a large number of new users came from All Things Alive, having already verified that "AI-powered apps" were a genuine need, and they were there to pay for and actually use the app.

Monthly income forecast jumped from 800,000 to over 2 million.

Su Niannian couldn't help but smile, "We didn't spend a single penny on promotion; it's all thanks to Dingsheng gifts."

"Next," Su Niannian patted the table, "It's Xiao Jiang's turn."

Jiang Yixin turned on her projector screen, then stood up and began to speak.

She meticulously categorized and statistically analyzed the user feedback tickets from the past two weeks. The table was densely packed, with each row representing a ticket, and the categories were marked with colors. It was very well organized, but there was too much information on one screen, making it difficult to see the key points.

She turned to the next page, her voice a little nervous. "The most frequently asked questions are concentrated on these points: 'How to connect to an existing system?', 'How to connect my own cloud account?', 'How to ensure backup and disaster recovery?', 'Can I bind my own domain name?'"

She paused, then added her own analysis, "These problems are concentrated among users in commercial scenarios—those who build membership systems, inventory management systems, and appointment scheduling systems. These people aren't making things for themselves; they're making them for their employees and customers."

Su Niannian added another layer of data, "The proportion of users in commercial scenarios has increased from 11% last month to 26%, and their payment rate and retention rate are much higher than those of ordinary tool users. Our fastest-growing user group happens to be the user group with the heaviest demand."

Su Niannian stood up, walked to the whiteboard, and drew two straight lines, dividing the whiteboard into three layers.

First layer: Local tools. Calculators, currency converters, information display pages, browsers or mobile apps, can be used immediately upon opening, no backend required. Open Things can handle these without any problems, and the cost is the lowest.

The second layer: personal data. This includes budgeting, journaling, and habit tracking. Local storage is adequate, but data is lost when switching devices. Connecting to iCloud or the Android system requires some developer knowledge. Occasional user complaints don't affect usability.

The third layer: a multi-user Production System. This includes membership management, inventory management, scheduling, and order management. It must have a database, Server, multi-device synchronization, and the ability for multiple users to operate simultaneously without generating corrupted data.

She drew another horizontal line below the third layer and wrote four words: Deploy the chasm.

" With Open Things's current deployment, when a user clicks 'Publish,' the app goes live and runs under our cloud account, shared by multiple tenants. Most users don't know or care about what's underneath."

95

Su Niannian put down her pen and turned around.

"But this architecture actually has problems. Last week, Qingyue Cloud experienced a regional fluctuation, and all the apps running on it went offline for a few minutes. The number of messages in the feedback group was so high that we couldn't reply to them all. One user who booked a nail salon appointment sent seventeen voice messages in a row."

"Currently, all user data runs on the same database cluster. Logical isolation is not physical isolation, which poses a risk of data access violations occurring simultaneously."

"Furthermore, customers have no control: they cannot choose the region, configure the domain name, or view the monitoring data."

Su Niannian glanced at the three people attending the meeting.

"For users who just want to play around, the current solution is sufficient. But those who actually use it for business need their own databases, their own domains, and their own backups. These things shouldn't be the responsibility of Open Things, and we don't want to be either, since we're not a cloud service provider."

The meeting room fell silent for a moment.

Lu Mingzhou broke the silence.

"Think about it the other way around."

He stood up and walked to the whiteboard, taking a calculator in his left hand.

"Currently, Open Things deploys all users' apps under their own cloud accounts, and all costs are borne by us. If we were to create an 'independent deployment' function, allowing users to bind their own cloud accounts and deploy their apps to their own cloud with one click, or help customers without cloud accounts create new accounts directly, Open Things would only be responsible for generating the code and the deployment process; users would pay for the underlying resources themselves, allowing them to choose according to their budget."

"Then we take a cut from the middle," Lu Mingzhou said, pressing the calculator a few times. "And it's not just a commission. Before, we negotiated a 66% annual discount with Azure Peak; we were the clients, buying their services. Now it's reversed. We have a large user base, and those users need cloud resources; we help the cloud vendors acquire customers. That changes the dynamic."

He wrote a set of numbers on the whiteboard:

With 200,000 monthly active users, (assuming) 5% have independent deployment needs, that's 10,000 users. Average monthly cloud consumption is 100 yuan per user, resulting in a monthly revenue of 100 yuan.

150,000. Get a 15% commission from Open Things = 150,000/month.

"The numbers aren't large right now," he patted the whiteboard, "But this growth rate is perfectly synchronized with the growth of Open Things's users, and the marginal cost is zero."

Su Niannian added, "Moreover, after independent deployment, the user's system architecture, technology stack, and update process are all generated by Open Things. Subsequent iterations and maintenance will still require them to come back to us. The deeper the usage, the more indispensable it becomes. These are all areas where we have bargaining power with cloud service providers."

Han Luyi glanced at the three-layer diagram on the whiteboard, then at Lu Mingzhou's numbers.

I'd like to have a chat with Liang Yu.

After the meeting, everyone else left the conference room, but Su Niannian kept Jiang Yixin behind.

Su Niannian said, "The report just now showed that the data was well organized and the work orders were categorized in great detail."

Jiang Yixin just breathed a sigh of relief.

"But you're providing 'what the user said,' not 'what the user wants.'"

Jiang Yixin didn't quite understand.

Su Niannian flipped to the page of her previous support ticket. "The user asked, 'Can I bind my own domain name?' This isn't just a technical issue, nor is it a deployment issue. If you dig one layer deeper, what they're asking is, 'Does the thing I spent time and money on actually belong to me?'"

Jiang Yixin nodded as she took out her notebook and began to take notes.

"There's a fundamental skill in product development: users tell you the symptoms, you need to find the root cause. The symptom is 'I don't know where the data is,' the cause is 'I need Production System, I need data security and business continuity.' There's a model in product methodology called 'What They Say, What They Want, What They Really Need,' a three-layer structure. What users say is the surface layer, what they want is the middle layer, and what they truly need is the bottom layer. Most product managers stop at the first layer, but that's far from enough. There's a book called Jobs to Be Done (JTBD); you can go back and read it, it will help you understand this model."

Jiang Yixin drew an iceberg in her notebook, wrote "Say" on the surface, labeled "Want" and "Need" on the bottom, and wrote the book title next to it.

The two left the meeting room and went to the break room to get water.

"President Su, how did you become a product manager at Beaver?"

Su Niannian poured herself a cup of hot water to make tea, and thought for a moment, "When I was in graduate school, I interviewed for a product management position. I like that – listening to people talk and then guessing the part they didn't say."

She blew on the steam rising from the rim of the cup and didn't say anything more.

7 PM, a private room at a home-style restaurant near Qiantan.

This is a long-term business banquet venue booked by Shen Congyun. It's not a big place, but it's quiet.

Han Luyi arrived first, Liang Yu seemed much more relaxed when he pushed open the door than the last time they met. Last time, when he ate with Xu Feng at a private restaurant in the alley, Liang Yu was somewhat tense. Nearly two months had passed, and the surge in Open Things's data gave him confidence. He was one of the first to bet on Open Things, Qingyue Cloud backend showed that the consumption of resources related to Open Things was skyrocketing.

"Boss Han, you're lucky," Liang Yu said with a laugh as he sat down. "Dingsheng brought this on themselves, giving you a massive influx of traffic."

Han Luyi exchanged a few pleasantries with him and then ordered the food.

After three rounds of tea and five courses of food, the tea was served.

Han Luyi refilled Liang Yu's teacup. "Open Things Liang, all of Kaiwu's users' applications are now running under our own Qingyue Cloud accounts."

"I know, we can see it from behind the scenes," Liang Yu said, picking up a piece of food. "It's rising very quickly."

"But this model has a few problems; some users have different needs," Han Luyi said. "We're developing a standalone deployment feature that allows users to link their own cloud accounts and deploy applications directly to their own cloud."

Liang Yu paused for a moment with his chopsticks.

"Is your 'Open Things' primarily aimed at non-developers?"

"Yes, many users don't have their own cloud accounts yet, but they need data isolation. In this case, Open Things can help them register new accounts."

"You mean you currently have unregistered users, and you want to become Azure Peak acquire customers?"

"Mr. Liang understands quickly; that's exactly what I mean."

"----What level can it reach?"

"The initial number is several thousand to ten thousand. Each user's monthly consumption isn't much, but this growth is completely synchronized with Open Things's user growth, so the prospects are very good. The more users we have, the more users will have cloud needs."

Liang Yu remained silent, picked up his teacup, and took a sip. The number itself wasn't particularly large— Qingyue Cloud Liushui was worth hundreds of millions—but the narrative was quite novel.

"Cloud vendors are all looking for new growth stories now," Liang Yu said slowly. "AI application distribution drives cloud consumption"—this narrative is so novel, no one will dislike it. I can use it to ask the channel cooperation department for resources.

Han Luyi nodded. "The first batch can only accept Qingyue Cloud."

He paused for a moment.

"But once the volume increases, we'll definitely start taking on other clients."

Liang Yu said, "Boss Han, is it true that you have to hit someone with a stick after giving them a sweet treat?"

"Mr. Liang, those who start running first gain a significant advantage," Han Luyi explained.

Liang Yu laughed, "You're just making empty promises to me."

F

"What you can actually eat isn't just empty promises."

Liang Yu looked at him for a while, then shook his head. "Boss Han, you really don't seem like a tech person anymore."

The direction of action was quickly determined: Liang Yu went back to the channel cooperation department to evaluate the cooperation model: Open Things first produce an independent deployment technical solution and API documentation; the commission rate would be discussed later, the cooperation model would be tested first, and the execution teams of both parties would meet once within two weeks.

The two ate for an hour and a half. Although they didn't drink alcohol, they had a great time. Liang Yu drove away, Han Luyi got into Zhang Biao's car and went home.

Han Luyi got home close to nine o'clock.

After taking a shower, he leaned back on the sofa and started flipping through today's tech news.

"Geek Review," a reputable tech industry media outlet, published a major headline.

<Code Guardian or Code Harvester?? The Data Security Mystery of BugKiller>

Han Luyi glanced inside.

The article begins by explaining how BugKiller works, written in a very "rational and objective" manner: BugKiller needs to scan the user's complete source code to detect bugs; scanning means acquisition, acquisition means storage, and storage means it can be used for other purposes.

The story progresses layer by layer, with each paragraph hinting that BugKiller stole the client's code.

The final paragraph summarizes. "According to sources familiar with the matter, Source Code Technology is internally training its self-developed Large Model, 'Heavenly Craft,' but the source of the training data is questionable."

Han Luyi exited the article and glanced at the publication time.

2 PM today.

This article was already published while he was still in his office.

As soon as All Things Alive and collapsed, someone came to smear BugKiller.

Han Luyi opened his Sight and looked at the news article again.

This is no coincidence.

End of Chapter
Prev
NovelSummaries
Next

Explore the wiki

Han Luyi
Su Niannian
Jiang Yixin
Zhao Wenyuan

Comments

No comments yet.
AboutStatsAPIChangelogStatusContactPrivacyTermsDMCA
© 2026 RandomTranslator·v3.26·RSS