How ErrorLens Works

AI-powered root cause analysis in three simple steps

Step 1 — Upload your error logs

ErrorLens accepts error logs in any text-based format. Drag and drop files directly onto the upload zone, or paste your stack trace or log content into the text area. Supported formats include .log, .txt, .json, .csv, .xml, and any plain text output.

Large files are automatically handled — content up to 500,000 characters is processed in full. For larger files, ErrorLens intelligently preserves the beginning (where initialization errors appear) and end (where the most recent failures are) of your log.

Step 2 — AI analysis engine

Once you submit, ErrorLens passes your log content to our analysis engine powered by Claude — Anthropic's state-of-the-art AI. The engine performs several passes:

• Error extraction — identifies every distinct error type, exception, warning, and failure signal.
• Occurrence counting — counts exactly how many times each error appears.
• Severity classification — Critical, High, Medium, or Low based on context and impact.
• Root cause inference — traces the underlying cause from stack frames, error messages, and contextual clues.
• Code patch generation — produces before/after code fixes inferred from class names, method signatures, and stack traces.

Step 3 — Prioritized results

Results are delivered as a prioritized list of errors, sorted by severity (Critical first) then by occurrence count. Each error card shows the root cause, numbered solution steps, a prevention tip, and side-by-side code diffs where applicable.

All analyses are saved to your team's history and remain searchable and filterable for the duration of your plan's retention period.