Architecture of an AI-based university KPI system: React.js, ASP.NET Core, and PostgreSQL

This article describes a layered architecture of a system designed to calculate university employees’ KPIs, detect anomalies, and provide recommendations. The technology stack includes React.js, ASP.NET Core, PostgreSQL, and a separate Python-based machine learning service. The ETL quality index (q_t), Isolation Forest (F1 = 0.81), and SHAP-based templates technically support the system’s decision support system (DSS) constraints. During load testing with 400 concurrent users, the system achieved a p95 response time of 198 ms.

Hali tarjima qilinmagan