Traces: Observe Your Agents and Then Improve Them

In the Haystack Enterprise Platform, Traces provide comprehensive visibility into the execution of production agents by capturing the complete lifecycle of pipeline runs, from inputs to outputs, without requiring additional observability tools. This feature records every run's hierarchical component tree, token usage, logs, failures, and metadata, offering an interactive waterfall view for detailed analysis. Users can quickly identify issues, investigate latency, and understand component behavior, allowing for immediate reproduction and testing of fixes. Traces also facilitate comparison between pipeline versions, highlighting changes in latency, token consumption, and outputs, thus empowering users to assess improvements and costs accurately. Data security is maintained with role-based access controls and encrypted storage, ensuring sensitive information remains secure on deepset-owned infrastructure, with the option to export traces as needed. Available now in the platform, Traces automatically generate upon deploying or redeploying pipelines, and users can explore its capabilities through documentation or demos tailored to their use cases.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.