Native tracing
Dekaplane ties together control-plane state, workflow runs, trace events, tool calls, and gateway behavior so operators can debug and govern AI workflows.
AI-assisted operations need a clear answer to a simple question: what happened? Without traces, teams are left reconstructing model calls, tool calls, and workflow behavior from scattered logs.
Dekaplane includes observability and tracing foundations so operators can connect requests, agent steps, tool execution, and control-plane configuration.
Tracing should follow the path from an operator request through workflow orchestration, model access, tool calls, and final output. Dekaplane surfaces that path as part of the platform.
A trace is most useful when it connects the control-plane decision to runtime behavior. That connection helps teams debug failures, review policy, and understand how an action was produced.
Dekaplane's documentation describes observability foundations including workflow runs, trace events, audit records, and telemetry-oriented integrations such as LangFuse and OpenTelemetry where configured.
Related reading
Current release
Dekaplane is currently available as a Free single-tenant self-hosted release with portal-based downloads, license materials, and installation guidance.
