Abstract: The system obtains KPIs of a cell of a telecommunication network. The system preprocesses the multiple KPIs by reducing noise of the multiple KPIs and normalizing an amplitude of a KPI among the multiple KPIs. The system identifies a performance issue of the network by executing multiple ML models based on the KPIs, where a first ML model is configured to identify a sudden performance degradation, where a second ML model is configured to identify an event leading to a change in a performance trend, where a third ML model is configured to identify a gradual change in the performance trend, and where a fourth ML model is configured to identify whether the cell of the network is nearing a capacity limitation of the cell. The system provides a report based on the identified performance issue.