n8n Workflow for Job Search and Job Applications (Free Template Outline, No LinkedIn Scraping)
Search for "n8n workflow for job application" or "n8n job search workflow" and most templates you'll find do two things: scrape job boards and auto-generate applications. Scraping breaks when sites change their HTML, and some sites' terms prohibit it. Auto-applying tends to produce generic applications. This outline keeps the useful part, deciding fast which jobs deserve your effort, and leaves the risky part out.
What the workflow does
| Step | n8n node | Output |
|---|---|---|
| 1. Capture the job | Form Trigger (n8n-hosted form) | Company, title, link and the full job description, pasted by you |
| 2. Load your profile | Code or Google Sheets | Master resume text, target titles, locations, salary floor, dealbreakers |
| 3. Match score | HTTP Request or AI Agent (JSON output) | match_pct, must-haves, matched and missing keywords, red flags |
| 4. Decide | Code / If | GO, STRETCH or SKIP from your thresholds, not the model's opinion |
| 5. Tailor (GO/STRETCH only) | HTTP Request or AI Agent | A 150–220 word cover note, 3–5 bullets, and "keywords you can't honestly claim" |
| 6. Log and remind | Google Sheets + Gmail + Schedule Trigger | A tracker row with a follow-up date, and a weekday email listing follow-ups due |
Why paste the job instead of scraping it
- It doesn't break. No HTML selectors to maintain when a job board redesigns.
- It works with any board. LinkedIn, Indeed, company career pages, referrals, email.
- It keeps your accounts safe. LinkedIn's User Agreement prohibits bots and other automated methods of accessing the service.
- The AI gets the full requirements, not a truncated card.
Free match-score prompt (system message)
You compare one job description to one resume. Use ONLY facts in the resume; never invent experience.
Return JSON: {"match_pct":0-100,"must_haves":[],"matched_keywords":[],"missing_keywords":[],"red_flags":[],"salary_posted":"text or null","why":"one line"}
Score conservatively. A must-have that is missing caps match_pct at 60.
JOB: {JOB_TEXT}
RESUME: {RESUME_TEXT}
Tracker columns (Google Sheets)
date, company, role, source, url, match_pct, decision, status, follow_up_date, missing_keywords, notes
Store follow_up_date as YYYY-MM-DD text so the reminder step is a simple comparison. See the tracker and reminder guide for the 4-node reminder workflow, and what to automate (and what not to) in a job search.
Build it yourself or import it
Everything above is enough to build this in n8n yourself. If you'd rather import it, the AI Job Search Autopilot is this workflow, finished: the form, scoring, thresholds, tailoring, tracker, reminder workflow, a 9-prompt pack for ChatGPT/Claude, and a setup guide.
Free: AI Prompt Pack for Wholesalers & Job Seekers
10 copy-paste prompts from our paid packs: seller lead triage, underwriting, objection replies, job fit scoring, honest cover notes and interview prep. Free download (PDF); Gumroad asks for your email so it can send you the file.
Get the free prompt packNo email? Download the PDF directly.
Want the full system?
The AI Job Search Autopilot scores each job against your real resume, drafts a tailored cover note and bullets (no invented facts), builds interview prep, logs everything to a tracker and reminds you to follow up. No auto-apply and no LinkedIn bots.
See the Job Autopilot: $29FAQ
Is there a free n8n workflow for job applications?
Yes. The outline on this page (form, match score, decision rule, tailoring, Google Sheets tracker and reminders) is free to build yourself. The n8n template library also has free job search templates; many of them scrape job boards.
Does this workflow apply to jobs automatically?
No. It prepares the decision and the drafts; you review and submit each application yourself.
Which AI model does it need?
Any chat model that can return JSON, called through n8n's OpenAI-compatible nodes or an HTTP Request node. You pay your provider for usage.