Some of our customers run their own online stores, and they ask us to ship orders directly to their buyers — the classic drop-shipping setup. It sounds simple, but it drags a surprisingly heavy problem along with it: shipping fees.

The problem: every package has a number, every number has a fee

Every package we send has a tracking number, and every tracking number carries its own shipping fee — the fee changes with destination, weight and season. When the month is quiet, we ship about 1,500 packages. At the year-end peak, that number climbs to nearly 10,000 packages in a single month.

At the end of every month, each drop-shipping customer asks the same question: "How much do I owe you for shipping?"

To answer it, someone had to sit down with two tables:

• Table 1 — the customer's order list: every tracking number we shipped for that customer
• Table 2 — the courier's fee list: every tracking number and its shipping fee

If a customer had 200 packages, that meant matching 200 tracking numbers against the fee list, line by line. Copy, search, paste, repeat. With a few such customers, a whole day could disappear into a spreadsheet — and the more numbers you match by hand, the easier it is to make a mistake.

The agent: two tables in, one statement out

So we built an agent. The workflow is almost embarrassingly simple:

• Export the two tables from our shipping system.
• Send both tables to the agent — nothing else, no special formatting required.
• The agent matches each of the customer's tracking numbers to its fee, checks the mismatches, and writes the settlement: package count, total fee, and a line-by-line list.
• We review it and send the statement to the customer.

That's it. One customer takes 1–2 minutes. What used to eat hours now finishes faster than a coffee break.

Where AI actually works here

To be honest, a simple Excel formula could do the matching too. So why an AI agent?

• The two tables never arrive in the same format. Courier exports change, columns get renamed, numbers come in different formats. The agent understands the messy reality instead of breaking on it.
• Edge cases get handled: returned packages, adjusted fees, special shipping lanes.
• The output is customer-ready — not a raw data join, but a clean statement with a summary on top.

And it scales. Whether the customer had 200 packages or 8,000, the agent doesn't care. That is exactly why we built it this way.

Another step on our AI road

We keep repeating the same idea in this blog: find the repetitive, boring, error-prone task, turn it into a tool, and share what works. Monthly shipping fee settlement was one of the most repetitive jobs in our office. Now it is a two-minute conversation with an agent.

And it directly benefits our customers. If you run a store and we drop-ship for you, your monthly shipping statement now comes faster and more accurate than ever.

Über unsere Fabrik (Deutsch)

Wir sind die „Yiwu Hongda Garment Factory" (义乌市弘达服装厂), eine Näherei in Yiwu, China, die sich seit 21 Jahren auf die Herstellung von Bekleidung spezialisiert hat.

• Bekleidung: Pyjamas, Hausbekleidung, Umstandsmode, Freizeitkleidung
• Taschen: Canvas-Tragetaschen, Umhängetaschen, Rucksäcke, Mom-Bags, Strandtaschen, Aufbewahrungstaschen
• Tierbekleidung: Hunde- und Katzenbekleidung, Kleider, Mäntel, saisonale und festliche Outfits

Unsere Dienstleistungen: Musterfertigung, OEM-Produktion und Fertigung unter Ihrer eigenen Marke (Private Label). Wir akzeptieren kleine wie große Bestellmengen, für den Inlands- wie den Exportmarkt.

Kontakt: WeChat / WhatsApp +86 189 5700 2020 ／ +86 167 5777 7733