Parse and process

Goal: Raw log messages are structured into queryable fields.

User adds processing rules to extract fields from unstructured logs. AI generates a complete pipeline from sample data. User previews results, refines patterns or iteratively guides the AI until the output is correct, then deploys with confidence.

Why Streams: No ingest pipeline expertise needed. No pipeline JSON. No Logstash. The AI generates patterns, the UI shows a live preview of the resulting documents, the expertise barrier is gone.