“Anomaly Detection and Prediction in Zabbix” – Aldin Osmanagić

This talk focuses on the importance of detecting anomalies and predicting future system behavior in modern IT environments to enhance service reliability and prevent downtime. The speaker, Aldin Osmanagic, discusses how Zabbix, an open-source monitoring platform, can provide enterprise-class observability without vendor lock-in. Key topics include baselining techniques, dynamic anomaly detection, prediction functions, and the comparison between machine learning and statistical approaches. Attendees will gain practical insights into setting up baselines and configuring anomaly detection and predictive triggers within Zabbix, showcasing the platform's capability to compete with proprietary solutions.