Predictive Medicine Demand
AI predicts next month's medicine demand with Pakistan seasonality factors
Overview
EliteHMS forecasts demand for each medicine using a weighted moving average (recent 3 months: 3:2:1 weight, blended 70/30 with historical average) enhanced by Pakistan-specific seasonality multipliers. Categories like Antibiotics peak in winter (×1.3) and GI medicines in monsoon (×1.4). Gemini AI adds Pakistan-specific market insights for the top 25 medicines. Accuracy is tracked monthly against actuals.
Standard Operating Workflow
Historical Data
System automatically uses 12 months of billing sales data per medicine. No manual input needed.
Run Forecast
Weighted moving average + seasonality factor applied. Prediction, safety stock (20%), and current stock computed.
Order Suggestion
Suggested order qty = Predicted + Safety Stock − Current Stock. Urgency: Critical (< 7 days stock), High, Medium, Low.
AI Insights
Gemini generates Pakistan-specific text insight per medicine: 'Augmentin demand peaks in winter Pakistan — order immediately.'
Create PO
Filter Critical + High urgency → Create Purchase Order → PO created with AI quantities. Pharmacist reviews before sending.
Key Operational Benefits
Real-World Healthcare Scenarios
December Preparation
October forecast: Azithromycin predicted 87 units for December (winter ×1.3). Current stock: 35. System suggests ordering 69 units by November 18. Pharmacist creates PO from forecast. December arrives — no stockout.