Most pharmacy software is designed by people who have never spent a day behind a pharmacy counter. The workflows are wrong. The screens are wrong. The data model does not match how pharmacists actually think about medicines, patients, and prescriptions.

We spent three months in working pharmacies before we started building Pharmaco. Here is what we found — and how it shaped every design decision we made.

The Prescription Lifecycle Is Not Linear

In a textbook, a prescription flows neatly: doctor writes it, patient brings it, pharmacist fills it, patient picks it up. In reality, prescriptions are partial, modified, refilled, disputed, and expired. Patients forget them. Doctors call to revise them. Insurance rejects them. A data model that treats a prescription as a simple document fails immediately in production.

Pharmaco models prescriptions as state machines with a full audit trail. Every transition — from received to verified to dispensed to refilled — is recorded with timestamp, actor, and reason. This gives pharmacists a complete history and gives compliance teams an audit trail without any extra effort.

Inventory at the Medicine Level

Pharmacy inventory is not like retail inventory. The same molecule can come in five different brands, three dosages, and two formulations. A patient might be on a generic that is temporarily out of stock, requiring a therapeutic substitution that the dispensing pharmacist needs to flag to the prescribing doctor. Our inventory model is built around active ingredients and therapeutic equivalence, not just SKUs.

The DGDA Compliance Layer

Regulatory reporting in Bangladesh requires specific record formats for controlled substances, expiry tracking, and purchase-to-dispense reconciliation. We built this as a separate compliance layer that runs alongside normal operations rather than forcing pharmacists to change their workflow to satisfy a report format. The software adapts to the regulation; the pharmacist does not have to.

Beta Results

After three months of beta testing with partner pharmacies, average prescription processing time dropped by 40%. Stock discrepancy incidents — where physical count did not match system count — dropped by 68%. Both numbers exceeded our targets. The pharmacists told us it felt like the software understood what they were doing. That was the goal.