Out of Stock System for Drugstore


Predictive analytics module


One of Unicsoft clients, a large pharmaceutical network came up with request for a collaboration regarding development and implementation of OOS (Out of Stock system). For that current moment, Client had over 2000 drugstores all around the country and a reporting module comprised of reports concerning revenues and daily turnover.


The main goal was to set up a proper ETL process and perform a predictive analytics module intended to predict absence of drugs & remedies alongside with BI-tool deployment for ad-hoc reports and predictive analytics. Subsequently, the request evolved into not only a analytics & reporting module, but a complex solution providing access for suppliers to let them see the picture considering sales & trends of their products in any point of sale.

Unicsoft started from setting a dedicated team comprised of data architects and engineers to establish and consequently maintain data migration process. Considering necessity of being able to do close to-real-time analytics as well as requirements of DB consistency and availability, a noSQL architecture was considered as the most suitable. Thence, a several data scientists were involved to manage data enrichment and munging process. Having this step completed, a predictive analytics module was implemented to calculate estimated risk of product absence and predicting sales trend. As a copestone, a BI module was added providing a visualization of predictive module outputs.



Client underlined well-organized and orchestrated delivery process from Unicsoft side, and, consequently, this case of collaboration led us to establishing a successful cooperation along with new projects for pharmaceutical industry.

TECHNOLOGY & TOOLS: noSQLPythonTableauXGBoostBig Data