LOW / CODEBlueprints
hrai

AI Resume Screener & Scorer

When resumes arrive via email, Claude AI automatically scores them against your job requirements and adds ranked results to Google Sheets with detailed reasoning. Your hiring team gets a Slack notification for top candidates.

Intermediate~25 minn8nMake.com 108 downloads 221 views

Setup Instructions

1. Get your Claude API key from console.anthropic.com and store it as a credential. 2. Create a Google Sheet with columns: Candidate Name, Email, Role Applied, Overall Score (0-100), Skills Match, Experience Match, Culture Fit Score, Strengths, Gaps, Reasoning, and Date Received. 3. Set up a Gmail trigger that watches for new emails matching a label or filter (e.g., subject contains "Application" or emails to careers@yourcompany.com). 4. Add a node to extract the resume attachment (PDF or DOCX) and convert it to text. Use a PDF parse node or HTTP request to a document parsing service. 5. Add an HTTP Request node to call the Claude API (POST https://api.anthropic.com/v1/messages) with model "claude-sonnet-4-20250514". In the system prompt, include the job requirements, required skills, and evaluation rubric. Ask Claude to return JSON with: candidateName, overallScore (0-100), skillsMatch (0-100), experienceMatch (0-100), cultureFit (0-100), strengths (array), gaps (array), reasoning (string). 6. Parse the Claude response and map fields to your Google Sheet columns. Append a new row for each resume. 7. Add a filter node: if overallScore > 80, send a Slack notification to your hiring channel with the candidate name, score, and top strengths. 8. Test by sending a sample resume email and verify the full pipeline — parsing, scoring, sheet update, and Slack alert.
Troubleshooting
**Resume text extraction fails for PDFs:** Some PDFs are image-based (scanned). Use an OCR service (like Google Vision API or a dedicated PDF-to-text node) before sending to Claude. Plain text and DOCX files parse more reliably. **Claude scores seem inconsistent across candidates:** Ensure your system prompt includes a detailed rubric with weighted criteria. Be specific about what constitutes a 90 vs. a 70. Include example scores for calibration. **Token limit exceeded for long resumes:** Set max_tokens to 2048 for the response. If the resume text itself is very long, truncate it to the most relevant sections (experience, skills, education) before sending to Claude — the full resume rarely exceeds 4000 tokens. **Slack alerts not firing for high scorers:** Verify the filter condition checks the parsed numeric score, not a string. Ensure the JSON parse step converts overallScore to a number before the comparison.

Need a custom version?

We can build a tailored automation workflow for your specific needs.

New blueprints weekly

Get notified when we publish new automation workflows.