[Logistics System] 1997: Verify tens of thousands of records in 5 seconds

Wu Chao Hui (JEFFI CHAO HUI WU)

Article Date: June 23, 2025, Monday, 2:12 PM

In 1997, I returned to my country to seek investment and visited a friend who was in the logistics business. When we talked about the management of their company, he suddenly complained about the reconciliation issues in the finance department: every month, just reconciling eight thousand bank statements with internal transactions took a full three weeks, with two finance staff working overtime, and they often made mistakes, which meant starting over whenever an error occurred.

After I finished listening, I asked, "How did you verify it?" He pointed to a thick stack of ledgers on the table and two Excel spreadsheets on the computer screen, saying, "One is the bank statement, and the other is the company ledger, manually checked line by line."

I glanced at the data and smiled, saying, "If it's just matching this kind of data, 5 seconds is enough." He was taken aback and replied, "Don't joke around, we are using Excel, but we still have to look at them one by one."

It must be added that it was Excel from 1997, with extremely primitive features, lacking all the advanced filters, conditional formatting, formula libraries, macro recording, and certainly any powerful tools like pivot tables, VLOOKUP, or Power Query that are available in today's versions. At that time, the only tools available were the most basic cells, formulas, and manual calculations.

I didn't say much; I simply asked them to open the spreadsheet, sat down, and spent fifteen minutes writing a simple reconciliation program. Then I pressed the Enter key, and the computer took only about 5 seconds to list all the unmatched records. The whole process was clear and straightforward.

The finance staff was stunned on the spot: "Is this done already?" "We always thought this kind of thing could only be done manually... and it was resolved in just 5 seconds?"

My friend excitedly slapped his thigh and said, "Mr. Wu, why don't we start a company to promote this system!"

I smiled and said, "This is just a small program, not worth starting a business." To me, this was merely a trivial issue, something I could solve easily.

But in my heart, I know that behind this so-called mini-program are two completely different ways of thinking: they are accustomed to using manpower to gradually fill the gaps in the system, while I am used to building systems with logic to fundamentally solve problems.

In my eyes, those three weeks of overtime were not a symbol of hard work, but a sign of a backward system. They relied on manpower, while I relied on the system. They used Excel for calculations, while I used programs for matching. They could not break free from inertia, while I changed their entire reconciliation process in just fifteen minutes.

In the end, they did not use my app. Years later, I finally understood that if I could complete the workload of both of them in three weeks in just 5 seconds, their positions would be at risk.

I remember that when I told this story to others, many just laughed it off. The capital market was also uninterested, thinking it was merely a "small tool" that couldn't do big business. But what they didn't see was that this capability is the prototype of future intelligent systems.

Today, many people can use Excel for reconciliation, even with the assistance of AI. But in 1997, that brief 5 seconds was a groundbreaking demonstration.