Helping pharmacists order the right stock

The client

A company that helped pharmacists order from their suppliers.

The challenge

Pharmacists bought from several suppliers, each with its own price list, and the same medicine is often sold both under a brand name and as a cheaper generic. Working out what to order, and from whom, took time every day. And with broadband not yet everywhere, many pharmacies couldn't rely on being online.

What we built

• Matching of branded and generic versions of the same product, so like could be compared with like
• Each pharmacy's usage history, read from its own labelling system, so suggestions were based on what had really been dispensed
• A central database of price lists from the different suppliers
• Suggested orders worked out automatically for the pharmacist to check, feeding a central ordering system
• A Windows application for the pharmacists to use day to day

Working offline

Each pharmacy kept its own copy of the product and price database, so the pharmacist could work without a connection. We designed our own update system to keep those copies current: it sent only what had changed, compressed, so updates were small enough for a dial-up line.

Reading data nobody had documented

The labelling system used in the pharmacies had no published way to get at its records. We worked out how its database was structured, so the dispensing history could be used to work out what each pharmacy needed.

Technical details

A Windows Forms client with a local database, kept up to date by our own differential update format: compressed patches generated centrally and applied to each pharmacy's local copy. Usage history came from reverse-engineering the labelling system's database structure. Branded and generic equivalents were mapped centrally and combined with multi-supplier price lists to produce suggested orders for the central ordering system.

• .NET
• Windows Forms