AI Engineering
Learn LLMs, RAG, agents, secure tools, and production AI through practical insurance examples.
Traces, agent runs, LLM/tool calls, and state transitions
A trace connects the operations serving one request or task. Each timed operation is a span, such as an LLM call, retrieval, or tool execution. An agent run may include several spans and persisted state transitions across multiple requests.
Latency, tokens, cost, errors, and service objectives
Latency metrics describe waiting time, including queue delay, time to first token, and total completion time. Token and cost metrics show resource usage. Error metrics classify failed, blocked, incomplete, and unknown-outcome runs.
Langfuse, LangSmith, OpenTelemetry, and Sentry
Langfuse and LangSmith provide AI-oriented trace and experiment views. OpenTelemetry standardizes telemetry instrumentation and export across services. Sentry provides application error and performance investigation capabilities; exact AI-specific…
A realistic production trace
Illustrative timings and identifiers:
Difficulty is an editorial estimate. Completion means you chose to mark a question as understood.