What happens when a 50-year-old bakery owner locks himself in his room over Chinese New Year and starts building an AI system with a tool called "Button"? You don’t get a story about tech enthusiasm — you get a real case of someone rediscovering his own capacity.
Ye Chengyu runs a chain of 15 bakeries in Wenzhou, with annual revenue over 20 million yuan. He was an early internet entrepreneur back in the PC era, built a food delivery website. But coding had long become a distant memory. This time, he used AI vibe coding — letting AI write most of the code while he guided it. When the system first ran, he felt a surge: "I can do this again."
That "reignition" wasn’t wishful thinking. It was a tangible sense of ability upgrade. And then he started applying AI to the most painful spots in his business.
Scenario one: data monitoring
With over 200 SKUs, he always thought the best-selling items made the most profit. But after using AI to analyze delivery data, he discovered the top-selling products had razor-thin margins — platform subsidies were creating a false boom. He adjusted promotions, cut the bundled deals. Sales volume dropped, but total gross profit rose. He finally started keeping the money he deserved.
Scenario two: meeting management
Before, meetings were chaos: data didn’t match, discussions wandered, introverts said nothing, and problems evaporated afterward. Now AI pre-checks data, flags inconsistencies before the meeting; during the meeting, it proactively calls on quiet team members to speak; afterward, it automatically summarizes everyone’s input. Everyone gets heard, every issue gets a follow-up.
Scenario three: execution tracking
Store managers used to write weekly reports that were just mindless logs — the boss got nothing useful. Now AI asks managers specific questions: "What’s this week’s bottleneck? Which channel is dropping? Why?" Managers just answer, no need to force a report. The boss gets precise information, not fluff. Ye says this was the smallest change, but managers felt it most directly.
Ye himself said: "Don’t overestimate what AI can change in the short term, but don’t underestimate the change from long-term persistence."
The real lesson? In the AI era, anyone can be reignited — not by willpower, but by finding one scenario where AI opens a boundary for you. Ye has been doing bakery for over a decade. AI didn’t replace him; it helped him discover he could do even more. The key is to find your scenario, and start from the most painful point.