Monitor for issues in your AI application

Hey Emre Elbeyoglu,
 
Reading traces one by one does not scale over time. That's where evaluators come in: they score traces for you, so the bad runs
are surfaced automatically. We call this monitoring
[https://langfuse.com/academy/monitoring?ref=onboarding_emails_v2_3_monitoring] and it represents the second step of the AI
Engineering Loop. [https://langfuse.com/academy/monitoring?ref=onboarding_emails_v2_3_monitoring]
 



 
Choose what to evaluate [https://langfuse.com/academy/evaluate/choosing-what-to-evaluate?ref=onboarding_emails_v2_3_monitoring]
Don't try to cover every failure mode up front. Pick the one failure or signal you've already seen with your own eyes or do error
analysis [https://langfuse.com/academy/monitoring/error-analysis?ref=onboarding_emails_v2_3_monitoring] on your traces to identify
a failure mode. You can also review our template gallery
[https://cloud.langfuse.com/project/~/evals?ref=onboarding_emails_v2_3_monitoring] to get inspiration.
 
Set up your first evaluator [https://langfuse.com/docs/evaluation/overview?ref=onboarding_emails_v2_3_monitoring]
Score every trace against one thing you care about. You can use LLM-as-a-judge
[https://langfuse.com/docs/evaluation/evaluation-methods/llm-as-a-judge?ref=onboarding_emails_v2_3_monitoring], custom code evals
[https://langfuse.com/docs/evaluation/evaluation-methods/code-evaluators?ref=onboarding_emails_v2_3_monitoring], or user feedback
[https://langfuse.com/docs/observability/features/user-feedback?ref=onboarding_emails_v2_3_monitoring], and track the trend on a
dashboard [https://langfuse.com/docs/metrics/features/custom-dashboards?ref=onboarding_emails_v2_3_monitoring] or use the scores
to filter on interesting traces.
 



 
 
Full detail in the evaluation docs [https://langfuse.com/docs/evaluation/overview?ref=onboarding_emails_v2_3_monitoring] and the
academy [https://langfuse.com/academy/evaluate/choosing-what-to-evaluate?ref=onboarding_emails_v2_3_monitoring].
 
Best,
Marc
 
Co-founder, Langfuse (part of Clickhouse)
Links: Roadmap [https://langfuse.com/docs/roadmap?ref=onboarding_emails_v2_3_monitoring], Docs
[https://langfuse.com/docs?ref=onboarding_emails_v2_3_monitoring], Talk to us
[https://langfuse.com/talk-to-us?ref=onboarding_emails_v2_3_monitoring], Example Project
[https://langfuse.com/docs/demo?ref=onboarding_emails_v2_3_monitoring]
 
You are receiving this email because you opted in to receive email from the list Onboarding Journey from Langfuse
Langfuse, Langfuse GmbH, Oranienburger Strasse 91, 10178, Berlin, Germany
Unsubscribe [https://app.loops.so/unsubscribe/cmv5q3x6a2x7501ui6r9t45s3/68a60e84a84c36110e706ac6d6b6c60f71b2047ecf182c5ce4d2cd8009b4f36c68aec531f49d2790ada4ef54bd28be66cf67e8b18e6fb8e925a6b363]

Accession

nse/langfuse/2026-10-12/3984

From

Marc Klingen · marc@updates.langfuse.com

Sent

12 October 2026 · 20:00 UTC