Applications increasingly use large language models (LLMs) to enhance AI capabilities, but observability is crucial to ensure their effectiveness and efficiency. Guardrails, a validation library for LLMs, has adopted OpenTelemetry to provide comprehensive observability support, allowing developers to monitor and verify LLM performance and accuracy. By enabling the tracking of various metrics such as execution times, validation success rates, and latency, developers can ensure reliable AI operations and quickly identify performance issues. OpenTelemetry, an open-source standard for observability, helps collect logs, metrics, and traces, which can be analyzed and visualized through tools like Grafana. This integration allows developers to maintain high-quality LLM applications and optimize their AI solutions by monitoring system health and performance continuously.