Error Tracking

Automatically detect, group, and investigate application errors with AI-powered analysis.

What Error Tracking Does

Bosca automatically captures JavaScript errors that occur in your web application — unhandled exceptions, failed promises, and runtime errors. Instead of presenting you with a raw list of thousands of individual error events, the system intelligently groups similar errors together so you can focus on distinct issues rather than individual occurrences.

Each error group gives you a consolidated view: what the error is, when it first appeared, when it last occurred, how many times it has happened, and a representative example with the full error details. This turns a flood of error notifications into a manageable list of actionable issues.

Error Groups

When an error occurs, the system examines three characteristics to determine which group it belongs to:

  • Application — Which application produced the error. Errors from different applications are always kept separate.
  • Error type — The kind of error (for example, TypeError, ReferenceError, or a custom error name).
  • Call path — The sequence of function calls that led to the error, with minor variations like line numbers removed so that the same bug is recognized even after code is redeployed.

These three characteristics are combined into a fingerprint — a unique identifier for that class of error. Every time the same kind of error occurs again (same app, same error type, same call path), it is added to the existing group rather than creating a new one.

The fingerprinting system is smart about code changes. If you redeploy your application and line numbers shift but the underlying code path is the same, errors will still be grouped together. Only genuinely different errors create new groups.

Each error group tracks:

FieldWhat It Tells You
Error type and messageWhat kind of error it is and the error message from the first occurrence
ApplicationWhich application produced the error
Fatal flagWhether the error crashed the application or was handled gracefully
First seen / Last seenWhen the error first appeared and when it most recently occurred
Event countTotal number of times this error has occurred
Sample eventA representative error event with full details for investigation
StatusCurrent workflow status: Open, Resolved, or Ignored
AssigneeThe team member responsible for investigating or fixing the error
AI summaryAI-generated root-cause analysis, if requested

Error Status Workflow

Every error group moves through a simple lifecycle that helps your team track progress on fixing issues:

StatusMeaningWhat Happens Next
OpenA new or recurring error that needs attention. All new error groups start here.Can be marked as Resolved (when fixed) or Ignored (when intentionally suppressed).
ResolvedThe error has been fixed and is no longer expected to occur.If a new event matches the same fingerprint, the group automatically reopens to alert you that the fix may not have worked.
IgnoredThe error is intentionally suppressed — it is a known issue you have decided not to fix.New occurrences are still counted, but the status does not change. You can reopen it at any time.
The automatic reopening of resolved errors is one of the most valuable features. If you mark an error as resolved and it comes back, you will know immediately rather than discovering it days later.

Assigning Errors to Team Members

Error groups can be assigned to specific team members to track who is responsible for investigating and fixing each issue. Assignment works independently of status — you can assign an error that is open, resolved, or ignored.

A typical workflow:

  1. A new error group appears with status Open
  2. A team lead reviews it and assigns it to the appropriate developer
  3. The developer investigates, deploys a fix, and marks the error as Resolved
  4. If the error reoccurs, it automatically reopens and the assigned developer is alerted

AI-Powered Root-Cause Analysis

Every error group has a one-click AI analysis feature. When you trigger an analysis, an AI agent examines the error and produces a human-readable summary of what went wrong and what to do about it.

The AI analysis considers:

  • The error type and error message
  • The stack trace showing where the error occurred
  • How frequently the error occurs and when it first appeared
  • The environmental context from the sample event (browser, page, device)

The result is a plain-language summary that explains the likely root cause and suggests next steps. This is especially helpful when:

  • You are triaging an unfamiliar error and need a quick understanding of what it means
  • You need to decide which team member to assign the error to
  • You want to assess the severity and urgency of an error without diving into code
AI analysis results are saved so you do not need to run the analysis again every time you look at the error. You can re-trigger the analysis at any time to get a fresh assessment — useful after the error accumulates more occurrences or after you have made code changes.

Filtering and Searching Errors

The error tracking interface provides several ways to find the errors you care about:

  • Filter by application — Focus on errors from a specific application when you have multiple apps
  • Filter by status — See only Open errors that need attention, or review Resolved errors to check they have not come back
  • Filter by severity — Show only fatal errors (those that crash the application) or non-fatal errors (those that were handled but still logged)
  • Text search — Search across error types and messages to find errors related to a specific feature or component

Filters can be combined. For example, you can search for all open, fatal errors in a specific application that mention "network" in the error message.

How Error Tracking Connects to Analytics

Error tracking is fully integrated with the rest of the analytics system. Error events flow through the same ingestion pipeline as all other events, which means:

  • Error events are enriched with geographic data, so you can see where errors are occurring
  • Error data is stored alongside all other analytics events and can be queried using the query system
  • You can build visualizations that display error trends, error rates by page, or error breakdowns by browser
  • Error metrics can be included on dashboards alongside traffic, engagement, and business metrics for a complete operational view
Add an error rate chart to your main operations dashboard. Combining error counts with traffic data lets you quickly distinguish between "more errors because more traffic" and "more errors because something is broken."