srbhr/Resume-Matcher · critical · HTTPException

Failed to improve resume. Please try again.

Error message

Failed to improve resume. Please try again.

What it means

HTTP 500 catch-all raised by improve_resume_endpoint for any unhandled exception during the resume improvement pipeline (AI calls, diff calculation, refinement, persistence). The real cause is only visible in the server log line 'Resume improvement failed: {e}'.

Source

Thrown at apps/backend/app/routers/resumes.py:1523

                # Diff metadata
                diff_summary=diff_summary,
                detailed_changes=detailed_changes,
                refinement_stats=refinement_stats,
                ats_score=_build_ats_score(
                    improved_data,
                    job_keywords,
                    refinement_result,
                    refinement_successful,
                ),
                warnings=response_warnings,
                refinement_attempted=refinement_attempted,
                refinement_successful=refinement_successful,
            ),
        )

    except Exception as e:
        logger.error(f"Resume improvement failed: {e}")
        raise HTTPException(
            status_code=500,
            detail="Failed to improve resume. Please try again.",
        )


@router.patch("/{resume_id}", response_model=ResumeFetchResponse)
async def update_resume_endpoint(
    resume_id: str, resume_data: ResumeData
) -> ResumeFetchResponse:
    """Update a resume with new structured data."""
    existing = await db.get_resume(resume_id)
    if not existing:
        raise HTTPException(status_code=404, detail="Resume not found")

    updated_data = resume_data.model_dump()
    updated_content = json.dumps(updated_data, indent=2)

    updated = await db.update_resume(

View on GitHub (pinned to 116f9cc3b0)

Solutions

1. Check backend logs for the 'Resume improvement failed: {e}' line to find the root cause
2. Retry the request — transient AI provider failures often resolve on retry
3. Verify AI provider credentials, quotas, and network connectivity from the backend
4. Reduce input size or retry later if the provider is rate-limited or down

Example fix

// server log shows: Resume improvement failed: RateLimitError: 429
// after: add backoff around the AI call in the pipeline
for attempt in range(3):
    try:
        return await call_llm(prompt)
    except RateLimitError:
        await asyncio.sleep(2 ** attempt)

async function improveWithRetry(req, attempts = 3) {
  for (let i = 0; i < attempts; i++) {
    try { return await improveResume(req); }
    catch (e) {
      if (e.status === 500 && i < attempts - 1) { await sleep(1000 * 2 ** i); continue; }
      if (e.status === 500) logServerSideHint('Check backend logs: Resume improvement failed');
      throw e;
    }
  }
}

When it happens

Trigger: AI provider timeout/rate limit or malformed response, exception in diff computation, refinement step failing, database write errors, or a bug in any pipeline stage not caught earlier.

Common situations: LLM API key invalid/quota exhausted, upstream AI service outage, oversized resume inputs, network egress blocked from the backend, or a code regression after a deploy.

Related errors

• Failed to load resume (status ${res.status}).
• Failed to delete resume (status ${res.status}): ${text}
• Failed to update cover letter (status ${res.status}): ${text
• Failed to update outreach message (status ${res.status}): ${
• Failed to rename resume (status ${res.status}): ${text}

AI-assisted analysis of srbhr/Resume-Matcher@116f9cc3b0 (2026-08-28). Data as JSON: /api/errors/4256fb47fc6f971d. Report an issue: GitHub.