Project

Log analysis using Large Language Models

Objective

Logs, often unstructured or semi-structured, contain vital information about system operations and user behavior. Analyzing these logs helps organizations identify errors, detect anomalies, and respond to security threats in real time. The advent of LLMs that are trained on vast amounts of textual data, enables them to understand complex language patterns and context, thus enhancing their capability to analyze logs accurately and efficiently. In this project, we evaluate the use of Large language models for understanding logs, and evaluate their cost and performance tradeoff.

Outcome

Paper at a conference