Job applicant CVs rated by AI, logged and sent to HR

Job applicant CVs rated by AI, logged and sent to HR

An n8n workflow for a job application form. It reads each applicant's PDF CV, has Gemini rate it against the job written in its prompt, adds a row to a Google Sheet, emails HR the rating and sends the applicant a receipt.

What is inside

• workflow.json: the workflow as n8n exports it, 7 nodes. All seven run, and there are no sticky notes. It holds no credential.
• README.md: these notes.

What we changed from the original copy

• The prompt in Using AI Analysis & Rating held two tokens that nothing filled in, ${output_language} and ${resume}. The model would have read them as literal text. The first is now the sentence "Write all of your output text in English." The second line now says the resume comes in the next message, because the node already sends the CV text as its own message.
• Google Gemini Chat Model had no model set, so it ran n8n's default for its version, models/gemini-2.5-flash. Google's deprecation page, read on 12 November 2026, says that name routes to gemini-3.5-flash-lite from 1 December 2026. The node now names models/gemini-3.8-flash, which Google's models page lists as stable. This is a plain chain with no tools.
• Both Gmail nodes are version 2.1, whose Email Type defaults to HTML and which adds a line saying the mail was sent with n8n. Both now send plain text with no such line, so the line breaks in the messages hold.
• Inform HR New CV Received sent to a named person's address. It now holds <YOUR_HR_EMAIL_ADDRESS>. The applicant's confirmation was signed with a company name, and it now ends with <YOUR_COMPANY_NAME>.
• The credential links are gone. The document and the tab in Candidate Lists are empty, so you pick your own.
• The file's name is ours.
• Nothing else. The form, the PDF step, the rules of the prompt and the job text are as the file had them.

Mindket read this copy against n8n's source and Google's pages. It was not run on a live n8n.

Before you start

• An n8n instance, on n8n Cloud or your own server.
• A Google account that can edit a sheet and send mail through Gmail.
• A Google AI Studio API key for the Gemini credential.
• A Google Sheet with one tab that has the columns CV, Full Name, E-mail, Expectation, Linkedin and AI Rating.
• The email address that should get the ratings, and your company name.

Set it up

1. In n8n, create a workflow, open its menu and choose Import from File. Pick workflow.json.
2. Add a Google Gemini (PaLM) credential with your AI Studio key and select it in Google Gemini Chat Model.
3. Add a Google Sheets credential and a Gmail credential. Select them in Candidate Lists, Inform HR New CV Received and Confirmation of CV Submission.
4. In Candidate Lists, pick your document and your tab.
5. In Inform HR New CV Received, replace <YOUR_HR_EMAIL_ADDRESS> with the address that gets the ratings. In Confirmation of CV Submission, replace <YOUR_COMPANY_NAME> with your company name.
6. Edit the job in the prompt of Using AI Analysis & Rating. It says "Software Engineer", and the form's title in Application Form says the same.
7. Open the form from its test address, submit a PDF CV of your own and check the sheet and both inboxes. Publish the workflow when you want a lasting form address.

What happens in a run

1. Application Form shows a form titled "Application for Software Engineer Position" with five required fields: Full Name, E-mail, Expectation (its hint reads 2000-3000$), Linkedin and a PDF upload, Your Resume/CV.
2. Convert Binary to Json extracts the text of the PDF from the binary field Your_Resume_CV.
3. Using AI Analysis & Rating sends the model the rules as the system message and the CV text as the user message. The rules ask for a score from 1 to 10 with the reasons, then a recommendation on an interview, in plain text and in no more than 75 words, for the job "Software Engineer".
4. Candidate Lists appends a row with the CV's file name, the name, the email, the expectation, the LinkedIn address and the rating under AI Rating.
5. Inform HR New CV Received emails the address you set, with the candidate's details and the rating.
6. Confirmation of CV Submission emails the applicant that the CV arrived and will be reviewed.

What to change

• The job and the rules of the rating: the prompt in Using AI Analysis & Rating.
• The language of the rating: the sentence about language in that prompt.
• The form's questions: Application Form. If you rename the CV field, change Your_Resume_CV in Convert Binary to Json too, since n8n builds that name from the label.
• The two emails: Inform HR New CV Received and Confirmation of CV Submission.
• The model: Google Gemini Chat Model.

Good to know

• The sheet's CV column holds the file's name, not the file.
• A PDF that is only a scan has no text for the model to read.
• The rating is a machine's first read of a CV. Open the CV before you decide anything about a person.
• The form has no sign-in. Anyone with its address can submit it, and each submission sends two emails and one model call.
• Data: the form collects names, emails, LinkedIn addresses and CVs. The CV text goes to Google, and the details go to your sheet and your inboxes. Check the rules on handling applicant data and on automated screening where you hire.
• Cost: one model call per applicant, billed by Google under your plan.