<![CDATA[The Purchasing Division of a pharmaceutical company plays a strategic role in inventory control and distribution fulfillment across retail outlets. Previously, the Key Performance Indicator (KPI) reporting system at PT. X was conducted manually using spreadsheets, requiring up to two working days and being prone to calculation errors that affected employee incentive payments. This study aims to design and implement a web-based KPI reporting system using Python, the Django framework, SQL Server with stored procedures, and the OpenPyXL library. The system was developed using the Waterfall methodology with a modified prototyping approach. Eighteen KPI indicators were classified into four categories: procurement efficiency, inventory and expiration management, distribution planning, and data compliance and negotiation. Data processing employed an SQL Server stored-procedure-based ETL pipeline integrated with Pandas and NumPy. The system reduced report generation time from two working days to less than one hour while eliminating incentive calculation errors.Keywords: Key performance indicator; Reporting automation; ETL pipeline; Purchasing; information systems.AbstrakDivisi purchasing pada perusahaan farmasi menjalankan peran strategis dalam pengendalian stok dan pemenuhan distribusi ke jaringan outlet. Sistem pelaporan Key Performance Indicator (KPI) PT. X sebelumnya dilakukan secara manual menggunakan spreadsheet dengan dua hari kerja dan rentan terhadap kesalahan kalkulasi yang berimplikasi pada pembayaran insentif karyawan. Penelitian ini bertujuan merancang dan mengimplementasikan sistem pelaporan KPI berbasis web menggunakan Python dan framework Django, SQL server dengan stored procedure, dan library openyxl. Sistem dikembangkan menggunakan metode waterfall dan dimodifikasi dengan pendekatan prototipe. 18 indikator KPI dikelompokkan kedalam empat kategori: efisiensi pengadaan (KPI 1), manajemen stok, dan kedaluwarsa (KPI 2), perencanaan distribusi (KPI 3), serta kepatuhan data dan negosiasi (KPI 4). Pengolahan data dilakukan melalui pipeline ETL berbasis SQL Server Stored Procedure dengan mengintegrasikan data heterogen, kemudian ditransformasi menggunakan pandas dan NumPy. Sistem berhasil mereduksi waktu penyusunan laporan dari rata-rata dua hari kerja menjadi kurang dari satu jam dan mengeliminasi kesalahan kalkulasi insentif. ]]>
