In early September, I wrote from a hotel in Bogotá about the first part of a business trip.
I said I would write the second part after returning home, including the work AI could not do for me. That part might be more important.
Now I was back. The expense report was submitted, the trip presentation delivered and the customer visit completed.
Here is the rest of that account.
Materials need to lead to a finished output
Start with expenses.
I recorded the first expense in Bogotá on August 30. On September 18, a formatted set of reimbursement documents came out of my computer.
The records covered nineteen days, three countries and five currencies.
During the trip, my repeated input was simple: after a purchase, take a photo of the receipt and put it into the working conversation.
The tool organized the receipts into an expense table with dates, amounts and purposes. It produced an index matching each image to its entry, then exported an A4 PDF ready for printing.
That brought scattered inputs through to a deliverable I could submit.
Anyone who travels for export sales knows the alternative: mixed currencies, cash and card payments, faded receipts and trying to remember what a transaction was for weeks later. Some claims end up unclear; others are never submitted.
What helped in this workflow was the evidence left for checking. There were backups before changes and notes showing whether an amount came from the actual card charge or a cash conversion, including the exchange-rate basis.
Photographing receipts reduced reconstruction and repeated entry. Before submission, amounts, receipt matches, duplicates and conversion rules still had to be checked against company finance requirements.
Meeting preparation happened before the meeting
A Colombian customer visited our factory in China. It was an established distributor with many customers and several represented brands, but little coverage of industrial Ethernet switches.
I had been waiting for that meeting for a long time.
Beforehand, I gave the assistant the customer's existing archive, several agency agreements from other markets, our partner-cooperation process and the questions this type of customer usually cares about.
It combined the materials into a meeting plan and agenda: discussion order, decisions to clarify in person and documents to have ready.
I reviewed the draft, made changes and used it.
The discussion and factory tour took more than three hours. We agreed on a framework for the agency relationship, with selection, samples and qualification documents still to follow.
In the past, preparing for such a meeting could mean an evening of searching through materials and anticipating questions.
That preparation did not negotiate for me. It helped me arrive with a more coherent plan.
The trip presentation came together during my commute
After returning, I needed to report on the trip.
Previously, that meant organizing information on dozens of customers, selecting and classifying hundreds of photos and assembling slides. Formatting alone could consume much of a day.
On this occasion, I provided customer information from the collected business cards, visit summaries and the trip photos while traveling from home to the office.
By the time I arrived, a presentation draft was ready.
The photos had been grouped by customer and setting rather than simply dropped into a sequence. I reviewed the result, changed some wording and used it for the report.
In this particular task, part of the commute became preparation time. That outcome depended on the materials and setup already available.
Keep the next action in a record
Expenses, agendas and slides were individual tasks. What made the month easier was connecting some of the work around them.
I already had customer materials and an email connection.
One side was relatively static: archives for more than forty target customers, covering their business, scale, represented brands, contacts and relationship history.
The other was dynamic: the development log, showing each contact round, sending date, reply and next action.
Previously, I had to remember to update one side after acting on the other. Busy periods created gaps.
I tried to link the two, update records after communication and compare them against original messages to identify omissions.
After meeting people at the exhibition and making further visits, I could more clearly see who needed which message, whose reply needed attention, and which account was waiting for samples or a quotation.
In my experience, an opportunity can be lost through an ordinary process failure: a forgotten follow-up or a late reply. Product and price still matter, but memory alone is a fragile way to keep every line of communication going.
What I valued was relying less on memory and more on records I could check.
One customer I met at the exhibition provides an example. I visited afterward, we selected products together and a quotation went out the same day. Less than a week had passed across the sequence.
The useful part was having the next action recorded at each point instead of leaving everything to recollection.
Two things I still had to do myself
The tools helped with more of the preparation, but two parts of the customer relationship remained mine.
During the visit, a conversation turned to the moon. The customer described what it meant in his culture. I talked about the Chinese Mid-Autumn Festival, the story of Chang'e and the association with family reunion.
He was interested and asked about mooncakes.
After taking him back to the hotel, I kept thinking about that detail. I arranged two flavors and had them delivered the next day, as I had promised. He appreciated the gesture and shared it with friends.
The useful action came from paying attention to the person in front of me. A line in a record saying “interested in mooncakes” would not have captured the whole situation.
The second part was business judgment.
Before the visit, the research pointed toward industrial switches as the opportunity. In face-to-face conversations, however, I noticed hesitation about adding a new category.
That hesitation was not explicit in the archive. It came through in pauses, tone and the way existing competitors and partners were discussed.
I changed the approach. Instead of pushing our strongest industrial line as the centerpiece, I positioned an offer to complement existing partners and fill gaps without directly conflicting with the brands already represented.
The relationship and subsequent discussion moved more smoothly.
Public information and model analysis can help us understand a customer. Actual intent still needs to be established through conversation.
AI helps organize and remind. People check facts and make decisions.
The distinction became clearer over that month: tools could support preparation, while direction, commitments and the human relationship remained my responsibility.
What it cost
This was not a free setup. I paid for a higher-tier subscription and additional usage credits.
I considered more than the subscription bill: time spent on administrative work, missed follow-ups and expenses I might otherwise fail to reconstruct.
Those trade-offs were specific to my workload. They were not a published price or a guaranteed return on purchasing an AI tool.
For me, comparing costs meant considering what was being missed as well as what was being spent.
One person, one month of work.
Before leaving for Bogotá, I described the setup as an experienced practitioner with an additional tool to amplify their effort.
After returning, I would add this: the point is to free attention for the work that still needs your judgment and presence.
This is an account of one business trip. Meeting progress, customer reactions and time spent depended on the available materials, tools, human effort and business conditions. It is not a promise that following the same steps will reproduce the same outcome.