Use n8n to match jobs against a resume

The builder vika2210 created a job-matching agent that uses a resume as context instead of relying only on a requested job title. A Google Drive folder receives the CV. The workflow downloads the PDF, extracts and splits its text, creates embeddings and stores the profile in Pinecone.

A message to the agent supplies the requested role and location. The AI Agent combines that request with the candidate profile and outputs structured search intent. Bright Data retrieves Indeed listings, while the workflow saves results in Google Sheets and emails the candidate. The post includes workflow screenshots, a demonstration link and a GitHub JSON artifact.

The implementation required resolving specific integration problems: keeping resume context consistent, making the agent return JSON instead of free text and mapping snapshot results into clean spreadsheet rows. The model roles were divided between embeddings, agent reasoning and job-description summaries. The demonstrated result is a personalized list, not submitted applications or an employment outcome. No candidate approval checkpoint before the email delivery is described.

Prompt
Original prompt

Source: dev.to · vika2210