n8n Workflow: Evidence-Based Professor Emails with GPT-4.1

Turn CV + motivation + professor research into a personalized email draft (with evidence checking)

This n8n workflow automatically creates an evidence-based outreach email: it reads a student’s CV and motivation letter PDFs, extracts research details from an official professor profile page, and uses GPT-4.1 to produce (and quality-check) a personalized email—then saves it as an unsent Gmail draft with both documents attached.

What this workflow does

• Ingests your submission: receives student details, a professor profile URL, and Google Drive links for the student’s CV and motivation letter via an n8n Form.
• Validates inputs & downloads PDFs: normalizes required fields, downloads both PDFs from Google Drive, and extracts their text content.
• Builds a student evidence profile: uses OpenAI (GPT-4.1) to structure evidence from the CV and motivation letter, stopping early if required fields are missing.
• Extracts professor details from a web page: fetches the professor profile via HTTP, removes scripts/styles, and uses OpenAI to extract structured research evidence.
• Scores research-fit & finds connections: compares student vs. professor evidence and outputs a 0–10 research-fit score plus the strongest documented connections.
• Writes and validates the outreach email: generates a personalized email (subject + body) and runs a second OpenAI validation pass to flag hallucinations or weak personalization.
• Creates a Gmail draft only when approved: if the professor email is valid and the email passes validation, it creates an unsent Gmail draft addressed to the professor, attaching the CV and motivation letter.

Use cases

• Automating research-group outreach for graduate admissions or collaboration proposals.
• Generating consistent, evidence-based professor emails while reducing “hallucination” risk.
• Building an n8n pipeline that drafts outreach emails and documents for later human review.

Technical details

• Integrations: OpenAI (GPT-4.1), Google Drive (PDF download), Gmail (draft creation), HTTP fetch of professor profile pages.
• Core node logic: if, set, code, html (HTML cleaning/processing), merge, gmail workflow step(s).