Datadog's log processing pipelines can automatically enrich logs with attributes that categorize them as they are ingested, enabling efficient analysis of applications and infrastructure at lower cost. Logs can be categorized based on URL patterns, response duration, or standard severity levels, allowing for the creation of custom categories, indexing policies, and metrics. These categories enable the exclusion of low-priority logs while archiving all enriched logs in long-term cloud storage. The use of log categories also facilitates the analysis of trends across all categories of logs, providing critical visibility into high-priority logs and long-term trends without the expense of indexing and retaining everything.