Robin enhanced observability for AI-enabled workflows in the fiddler-examples repository by introducing an OpenTelemetry context attribute specifically for large language model (LLM) spans. Using Python and leveraging OpenTelemetry, Robin improved the traceability of AI interactions by exposing LLM context attributes directly in telemetry traces, which streamlines debugging and data analysis. Robin also addressed a syntax issue to ensure the new attribute was logged correctly, reinforcing the reliability of the data logging mechanism. This work demonstrated a thoughtful approach to telemetry design, focusing on robust integration of AI context within existing Python-based logging and monitoring infrastructure for more effective workflow analysis.