Viewing Gentrace v2 documentation
View Gentrace v1
Gentrace home page
Search...
⌘K
Getting started
Get started
Quickstart
Installation
Initialization
Error analysis
Overview
Gentrace Chat
Derivations
Views
Alerting
Tracing
Overview
interaction()
traced()
Instrumentations
Manual LLM instrumentation
Evaluation
Overview
Experiments
Datasets
Unit tests
Dataset tests
Integrations
AI SDK
LangChain
LangGraph
Mastra
Next.js
OpenAI Agents
OpenAI (JS)
OpenAI (Python)
Pydantic AI
Administration
Single sign-on (SSO)
SCIM
Roles
Self-hosted deployment
Reference
OpenTelemetry guide
OpenTelemetry API
SDK Errors
SDK Entities
Datasets
Experiments
Pipelines
Test Cases
Support
Dashboard
Viewing Gentrace v2 documentation
View Gentrace v1
Gentrace home page
Search...
⌘K
Ask AI
Support
Dashboard
Dashboard
Search...
Navigation
Error analysis
Overview
Documentation
API Reference
Documentation
API Reference
Error analysis
Overview
Copy page
An overview of error analysis with Gentrace.
Copy page
In Gentrace, error analysis refers to the process of discovering, monitoring, and alerting on errors that happen in both your production agent and in local / test time.
As a pre-requisite for error analysis, ensure you are receiving either:
traces
from your local or production agent
experiments
from your evaluation infrastructure
Then,
ask
Gentrace chat
to explore your traces and experiments
have chat create
derivations
, monitoring rules for errors
configure
alerting
to be notified of errors
Previous
Gentrace Chat
An agent to explore traces and create derivations.
Next
Assistant
Responses are generated using AI and may contain mistakes.