Home / Industries / Trading & B2B
Industry · Trading & B2B SMEs

For trading & B2B SMEs

Get the quotations, receipts and stock out of WhatsApp and spreadsheets, and into a system that does the counting for you.

01

The work is fine. The paperwork is the problem.

Trading firms and B2B SMEs in Hong Kong rarely need a prettier website. They need the quotations, invoices, receipts and stock counts to stop living in WhatsApp, email attachments and one very important spreadsheet that only one person understands.

02

What we build for trading & B2B

OCR

Documents into data

Receipts, invoices and certificates read automatically and pushed into your accounts — the work behind the receipt product and OCR service we built.

Systems

Internal tools that replace the spreadsheet

Quotations, orders, stock and reconciliation in one place, with the rules your business actually follows.

Stock

Inventory & fulfilment

Live stock, low-stock alerts, packing lists and tracking — including the physical fulfilment queues we maintain.

Finance

Reporting & reconciliation

Monthly P&L, supplier cost comparison and reconciliation against the bank, instead of a week of copy-paste.

Portal

Customer & supplier portals

Let customers check their own orders and suppliers submit their own data, so the phone stops being the interface.

Automation

Automations

The quiet jobs: reminders, exports, alerts and scheduled reports that used to be somebody’s Friday afternoon.

03

B2B work we have done

Receipt capture, a rental operations platform, and document recognition — all in production.

04

Situations we see every month

Each of these is one module, not a system replacement.

05

Why a trading firm hires us

We built receipt capture and reconciliation for a business that had to live with it every day.

We ship the boring half

Parsing is the easy part; the hard part is the exceptions, the retries and the audit trail. That is where the work goes.

Calibrated on your documents

A parser tuned on samples nobody uses is worthless. We test on your real paperwork and report the accuracy we measured.

One module at a time

Start with the process that hurts most, watch it pay for itself, then decide about the next one.

Cheap to keep running

The systems we maintain have to stay cheap to operate, so we do not hand you an architecture that needs a full-time engineer.

06

FAQ

We run everything in Excel. Do we have to throw it away?+
No. We usually keep the spreadsheet as the starting point, automate the part that hurts most, and only replace what is worth replacing.
Can it read our supplier's documents?+
Usually yes — we calibrate recognition against your real documents, because a parser tuned on samples nobody uses is worthless.
Do we have to move to the cloud?+
Not necessarily, but it is usually cheaper and safer than a machine in the back office. We size it to your volume, not to a sales target.
Can you start small?+
That is the way we prefer it. One module, in production, earning its keep — then the next.
How accurate is document recognition, honestly?+
It depends on your documents, which is why we test on yours and give you the measured number before you commit. Nothing beats checking against the original for the fields that matter.
Can it run inside our own systems rather than yours?+
Yes. It runs in your own cloud account with your own data, and we hand over access whenever you want it.

Which task eats your week?

Name the process that hurts most. We will tell you whether it is worth automating, honestly.

Talk to us →