Transactional

Error Tracking Overview

Capture, track, and resolve errors in your applications with automatic grouping and intelligent alerting.

What is Error Tracking?

Error tracking automatically captures, groups, and manages errors across your applications. Similar to tools like Sentry or Bugsnag, it provides complete visibility into production errors with rich context for debugging.

Key Capabilities

  • Automatic Grouping - Similar errors are grouped into issues using fingerprinting
  • Full Context - Stack traces, breadcrumbs, user info, and request data
  • Multi-Platform - Support for JavaScript, Node.js, React, Next.js, Python, and more
  • Smart Alerts - Get notified about new issues, regressions, and error spikes

Why Use Error Tracking?

Production debugging is hard. Without proper error tracking, you face:

  • Blind Spots - Errors happen silently without visibility
  • Lost Context - Users report issues without reproduction steps
  • Alert Fatigue - Raw error logs are noisy and hard to prioritize
  • Slow Resolution - Finding root causes takes hours instead of minutes

Error tracking solves these problems with intelligent error management.

Key Features

Automatic Error Grouping

Errors are grouped into issues using fingerprinting. One issue represents many individual error occurrences:

Issue: "TypeError: Cannot read property 'name' of undefined"
├── Occurrence 1: User A, 10:30 AM
├── Occurrence 2: User B, 10:31 AM
├── Occurrence 3: User A, 10:45 AM
└── ... 47 more occurrences

This reduces noise and helps you focus on unique problems.

Rich Context Capture

Every error includes the context you need to debug:

  • Stack Traces - Full call stack with source map support
  • Breadcrumbs - Timeline of user actions leading to the error
  • User Info - Who was affected
  • Request Data - URL, headers, and body
  • Tags - Custom metadata for filtering

Issue Lifecycle

Track errors through resolution:

StatusDescription
UnresolvedNew or recurring issue needing attention
ResolvedFixed and deployed
IgnoredKnown issue, intentionally suppressed

Resolved issues automatically reopen if they recur (regression detection).

Intelligent Alerting

Get notified about what matters:

  • New Issue - First occurrence of a unique error
  • Regression - A resolved issue reoccurs
  • Threshold - Error count exceeds limit in time window

Send alerts via email, Slack, or webhooks.

Integration with LLM Observability

Error tracking integrates seamlessly with LLM tracing. When errors occur in your AI pipelines:

  • Errors link to the trace where they occurred
  • You can see the LLM inputs/outputs that led to the error
  • Cost and performance data is preserved

Supported Platforms

PlatformAuto-captureSource Maps
JavaScript (Browser)Uncaught exceptions, unhandled rejectionsYes
Node.jsUncaught exceptions, unhandled rejectionsYes
ReactError boundariesYes
Next.jsServer + client errorsYes
PythonException handlersSymbolication
GoPanic recoveryStack traces

Getting Started

  1. Install the SDK - Add error tracking to your application
  2. Configure Capture - Enable automatic or manual error capture
  3. Set Up Alerts - Configure notifications for your team
  4. Monitor Dashboard - View and manage issues

Ready to get started? Check out the Quickstart Guide.

Next Steps