Artificial Intelligence
Agent observability with OpenTelemetry GenAI semconv in 2026
The OTel GenAI spec stabilizes attributes for LLMs, tools, and agents. Practical Python implementation with Anthropic + Grafana Tempo.
Tag
trazas
Artificial Intelligence
AI agent observability: what to instrument first
Agents that chain calls to models, tools and memory are hard to debug without instrumentation designed for them. After a long year running agents in production, I cover what to measure first, which standards are consolidating, and which costly mistakes are avoided by getting the traces right from the start.
Artificial Intelligence
LLM Observability: Traces, Costs, and Quality
LLM applications need three distinct observability planes: prompt and response traces for debugging hallucinations, per-token and per-feature cost tracking, and response quality evaluation. Mature tools like Langfuse, LangSmith, and Helicone cover all three planes with specific instrumentation.
Technology
Grafana Beyla: Auto-Instrumentation Without Touching Code
Grafana Beyla is an eBPF agent that automatically instruments existing applications without touching their code: it observes kernel syscalls and generates OpenTelemetry traces and RED metrics for services written in Go, Java, Python, Node, and Rust. It gives broad, immediate coverage, but it does not replace the manual SDK for business metrics and internal logic.