The field of anomaly detection and timeseries monitoring is experiencing significant advancements, driven by the need for efficient and scalable solutions. Researchers are exploring novel approaches to improve the performance and interpretability of anomaly detection algorithms, such as adapting existing methods for streaming data and developing new techniques that combine the strengths of multiple approaches. Additionally, there is a growing interest in optimizing timeseries monitoring systems to reduce operational costs and query latency. Noteworthy papers in this area include:

• 8 Years of Optimizing Apache Otava, which presents a significant speedup in computing change points, and
• Online Isolation Forest, which proposes a novel method for anomaly detection in streaming contexts.
• PIF: Anomaly detection via preference embedding, which introduces a new method that combines adaptive isolation methods with preference embedding, and
• Approximation-First Timeseries Monitoring Query At Scale, which presents PromSketch, an approximation-first query framework that achieves significant reductions in query latency and operational costs.