Traces, logs and metrics
See inside production systems: errors, latency and LLM traces, the difference between guessing and knowing.
Seeing inside systems: metrics, logs, traces, and now LLM traces. The practice of debugging production with evidence.
Who it's for
Teams who run production systems and need to answer 'why is it broken' quickly.
Keep exploring
Concepts in the glossary
28 in catalog
Observability is featured by
-
Langfuse
The open-source LLM engineering platform
LLM Ops -
Helicone
LLM observability in one line
LLM Ops -
LangSmith
Build trustworthy AI agents
LLM Ops -
Braintrust
Enterprise AI evaluation
LLM Ops -
promptfoo
Test your LLM app before you ship
LLM Ops -
Portkey
The AI gateway and guardrails
LLM Ops -
Arize Phoenix
Open-source AI observability
LLM Ops -
Sentry
Fix what crashes, fast
Service -
Axiom
Logging and observability without limits
Service -
PostHog
The single platform for product data
Analytics Tool -
LiteLLM
One interface for all LLM APIs
LLM Ops -
Grafana
Metrics, logs, traces unified
Service -
Better Stack
Uptime, logs and on-call together
Service -
Checkly
Code-first synthetic monitoring
Service -
Datadog
Enterprise observability
Service -
New Relic
Full-stack APM
Service -
Uptime Kuma
Self-hosted uptime monitoring
Tool -
DeepEval
The open-source LLM evaluation framework
LLM Ops -
Pydantic Logfire
The AI observability platform that sees your whole app
LLM Ops -
W&B Weave
Observability and continuous improvement for production agents
LLM Ops -
Agenta
The open-source workspace for your agents
LLM Ops -
VictoriaMetrics
Simple, Reliable, Efficient
Hosting Service -
UptimeRobot
The leading uptime monitoring service
Service -
Atlassian Statuspage
Build customer trust with status pages
Service -
Instatus
Get ready for downtime
Service -
LogRocket
Session replay for web apps
Analytics Tool -
HyperDX
Resolve production issues, fast
Service -
Weights & Biases
The AI developer platform
LLM Ops