Guo wanted a company that could keep working when the people who began it were no longer in the room.
He presented the proposal at a meeting with Chen, Ning, Yanchuan and two engineers from prospective partner firms. The new company, Henglan Systems, would develop and support industrial decision tools under contracts that separated advisory recommendations from operational control. It would own some software, license some methods and employ people whose job was to maintain systems after the original demonstration.
“Why form another company?” Chen asked.
“Because we cannot finance twenty customers as twenty independent experiments,” Guo said. “And because customers need to know who answers the phone in the third year.”
Yanchuan liked the second reason more than the first.
The draft organization chart looked orderly. Product development at the top, delivery teams beneath, a legal and compliance group to one side. At the bottom was a support desk staffed by three people who would be expected to understand factories in several countries.
“Who can stop a deployment?” Ning asked.
Guo pointed to the project director.
“What if the director is rewarded for launching it?”
The room went quiet.
They changed the governance proposal to include an independent technical sign-off for safety-relevant changes, documented authority for local operators and a process for reporting unresolved findings without having to obtain the sales department’s permission. The structure did not guarantee courage. It made it harder for courage to require breaking a rule.
Henglan’s first capital plan assumed revenue from the factory pilot, the Pune inspection service and several projects that had not yet been signed. Yanchuan asked the analyst to separate contracted income from interest expressed by potential customers. The forecast shrank.
“Investors understand pipelines,” Guo said.
“Then label the pipeline,” Yanchuan replied.
The revised accounts showed how long Henglan could pay staff if its next two customers declined. The answer was shorter than Guo had hoped.
Guo set the printout on the table. “If I show this version, they’ll ask why I want to finance us at all.”
“Then explain the contracts we have,” Ning said.
“And if that’s not enough?”
“Then we hire fewer people,” Chen replied. Nobody liked how quickly the room agreed. Each vacant position was work someone else would have to carry.
He reduced the first hiring plan and reserved funds for support obligations already promised. The company would not be large enough to boast about. It might be large enough to survive a difficult quarter without abandoning a customer.
An read the governance summary at Yanchuan’s request. She asked what would happen if an operator reported a danger and the company decided the complaint was bad for its reputation.
“There’s a path for an independent review,” he said.
“Who pays the person who performs it?”
The draft had not said.
Guo approved a separate budget for reviews and a rule against tying the reviewers’ pay to the sales result. Yanchuan did not consider the problem solved. He considered the decision recorded.
At the founding meeting a lawyer asked whether Yanchuan wanted equity. The proposed share would make him financially dependent on Henglan’s valuation and complicate his role as an outside reviewer. He took the offer home and discussed it with Chen and An.
“You can own a part of it,” An said, “but then you need to know which questions you’ll be less free to ask.”
He chose a bounded advisory contract rather than a founding share. Guo was disappointed.
“I thought you wanted to build something lasting,” he said.
“I do. It doesn’t have to belong to me.”
Months later Henglan hired its first dedicated support engineer. He asked for access to the incident logs before accepting the role. Ning smiled and showed him the records, including failed tests.
“This company is going to be difficult,” Guo said.
“Perhaps that is why it might last,” she answered.
In the family file there was still no resolution of Yansheng’s disappearance. A company could grow, buildings could be filled and new contracts signed while an old absence remained. Yanchuan no longer expected the world to stop moving until he received an answer.
He signed his first advisory agreement with Henglan and asked them to include an exit process: how another qualified person would continue the work if he left.
Guo complained that the contract already contained too many pages about failure.
“Then shorten the sales deck,” Yanchuan said. “But keep the exit plan in the contract. Someone will need it after we’re gone.”