AI Enabled Observability Explainers - We Actually Did Something With AI!

Overview

Discover how eBay's Observability team combines algorithms with LLMs to create predictable "Explainers" for telemetry signals, enhancing trace, metric, and log interpretation beyond basic AI implementations.

Syllabus

- Introduction to AI in Observability -- Overview of Observability in modern systems -- The role of AI in enhancing observability -- Introduction to eBay's Observability team and their approach - Fundamentals of Telemetry Signals -- Understanding traces, metrics, and logs -- Challenges in telemetry signal interpretation - Algorithms for Signal Interpretation -- Review of foundational algorithms used in telemetry -- Advanced techniques for signal processing -- Case studies of algorithm applications in the industry - Introduction to Large Language Models (LLMs) -- What are LLMs and how do they work? -- Capabilities and limitations of current LLMs - Integrating LLMs with Observability Systems -- Strategies for combining algorithms and LLMs -- How LLMs can enhance trace, metric, and log interpretation -- Real-world examples and use cases from eBay - Building Predictable Explainers -- Concept of "Explainers" in AI -- Steps to develop explainers for telemetry data -- Predictability and reliability in AI explanations - Evaluation and Improvement -- Techniques for evaluating explainer effectiveness -- Feedback loops and continuous improvement - Future of AI in Observability -- Emerging trends and technologies -- The future landscape and potential advancements - Practicum and Case Study Analysis -- Hands-on projects/assignments -- Analyze case studies from eBay and other industry leaders - Conclusion and Industry Insights -- Key takeaways from the course -- Discussions with industry experts and professionals

Taught by

Tags