LLM Observability for Laravel - trace every AI call with Langfuse

How much did your LLM calls cost yesterday? Which prompts are slow? Are your RAG answers actually...