Datadog on Data Science скачать в хорошем качестве

Datadog on Data Science

In this episode we'll visit the world of predictive analytics and machine learning and uncover how these cutting-edge technologies are transforming the way Datadog monitors and improves its services. We’ll focus our conversation on two key aspects: using advanced statistical methods for proactive monitoring and the strategic implementation of machine learning for algorithm enhancement. Join us live to learn and ask questions about the application of Bayesian methods in detecting incidents in third-party SaaS APIs, a crucial step towards ensuring uninterrupted service and customer satisfaction. We’ll also reveal how strategic planning and execution in machine learning have led to significant improvements in critical operational algorithms. Bringing their expertise to the table, Anne-Marie Tousch and Clement Tiennot from Datadog's data science team will share these insights, offering a rare glimpse into the challenges and triumphs of applying data science in a fast-paced tech environment. 00:00 - Introductions 04:12 - What does Data Science mean at Datadog? 06:42 - Using data science to improve third party monitoring 10:59 - How is data science different than software engineering? 23:18 - Grouping monitoring data (Time series, Tags & entities, Rare events, Large scale) 27:57 - Log Anomaly Detection (Clustering / Stateful streaming) 31:36 - Using Hierarchical Bayesian models to monitor third party services 37:12 - Finding signals in metrics (Lage scale time series screening) 39:24 - Bits AI and generative artificial intelligence 45:23 - The challenge of evaluation (rare events vs precision & recall) 49:21 - Example of time series anomaly 51:18 - How is all of this tied back to Watchdog? 54:01 - Takeaways 57:36 - Questions & Answers (and Tools to check out)