Speed up first pass hiring with image based resume screening. The flow turns a PDF resume into an image and checks job fit using a vision model. Great for HR teams that need fast, fair shortlists without reading every resume by hand.

Run the test, pull a PDF from Google Drive, convert it to a JPG through a PDF to image service, and resize it to keep processing fast. A Google Gemini vision model reads the image like a person would and returns a clear JSON result with is_qualified and reason. A decision step then moves only qualified candidates forward to stage two. By reading the file as an image, the flow avoids hidden prompts and bad text extraction, so results are more trustworthy.

Use a private PDF conversion server for privacy and connect your Google accounts in n8n. Expect shorter screening cycles and fewer manual checks. This is helpful for teams that store resumes in Google Drive, handle mixed layouts, or worry about resumes that try to trick AI tools.

• Manual test trigger to run the screening on demand
• Google Drive download pulls the resume PDF by file ID
• HTTP request converts the PDF to a JPG using a PDF to image API
• Image resize reduces dimensions to speed up AI processing
• Google Gemini vision model evaluates job fit from the image
• Structured Output Parser forces a JSON with is_qualified and reason
• If node routes only qualified candidates to stage two
• LLM chain sends the prompt and image to the model in a controlled way

• Reduce manual screening time from 15 minutes per resume to about 2 minutes
• Automate up to 70 percent of first pass candidate review
• Lower false positives by ignoring hidden prompts in PDFs
• Handle up to 5 times more resumes per hour with the same team
• Connect file storage and AI analysis in one clear flow

1. Import the template into n8n: Create a new workflow in n8n > Click the three dots menu > Select 'Import from File' > Choose the downloaded JSON file.
2. You'll need accounts with Google Drive, Google Gemini and Stirling PDF. See the Tools Required section above for links to create accounts with these services.
3. Open the Google Drive node in n8n. In the Credential to connect with dropdown, click Create new credential and follow the steps to connect your Google Drive account. Confirm the file ID is set or use the file picker to select a sample resume.
4. Open the Google Gemini Chat Model node. In the credentials dropdown, click Create new credential for Google Gemini PaLM API. Get your API key from the Google AI Studio API page and paste it into n8n. Keep the model as gemini 1.5 pro latest.
5. Open the PDF to Image HTTP Request node. If you host your own Stirling PDF, replace the URL with your server endpoint. Public servers may work for tests, but use a private instance for production. No credential is needed unless your server requires auth.
6. In the HTTP Request node, confirm Content Type is multipart form data and the fileInput field maps to the binary data from the Google Drive node. Set image format to JPG and single page output. Test this node and check that a JPG binary is created.
7. Open the Edit Image node and confirm Resize by percent is set to 75. Increase or decrease if output quality is unclear or processing is slow.
8. Open the Structured Output Parser node and confirm the JSON schema includes is_qualified and reason. Add more fields if you want extra details like seniority or skill match.
9. Open the Candidate Resume Analyser chain node. Edit the prompt to include your role summary and required skills. Make sure the image binary from the Edit Image node is connected and the language model and output parser are linked.
10. Open the If node named Should Proceed To Stage 2 and ensure it checks the is_qualified boolean equals true. If you later add a score, change the condition accordingly.
11. Click Test workflow to run end to end. Verify the Google Drive download, the JPG conversion, the resized image, the model output fields, and the final routing in the If node.
12. Troubleshooting: If Google Drive fails, check file permissions or use a shared link. If the HTTP node returns 415, verify the form field names and content type. If Gemini returns 401 or empty output, recheck the API key, model name, and image size.