CORTEXFLOW · FREE FOREVER

Automation

AI Job Hunt Automator

↓ Download workflow JSON

Import in n8n: Workflows → ⋯ → Import from file, then add your credentials. Free download with a free account.

How it works

Your personal job-hunting assistant that runs every morning while you sleep:

  1. STEP 1 — Fetch new job listings — a daily 9 AM schedule trigger fires the JSearch API (RapidAPI, aggregates Google Jobs) with your keywords and location, then normalizes every posting into title, company, location, URL, and description.
  2. STEP 2 — Dedupe + log to Google Sheets — the workflow reads the job URLs already in your sheet, keeps only genuinely new postings, and appends them with status NEW.
  3. STEP 3 — AI resume match scoring — an AI Agent compares each job against your resume and returns a structured score (0–100), a verdict (STRONG_MATCH → NO_MATCH), top strengths, and key gaps via a Structured Output Parser. The sheet flips to SCORED.
  4. STEP 4 — Cover letters for top matches — jobs scoring at or above your threshold get a tailored cover letter (max 250 words, referencing your real achievements) written by a second AI Agent, saved to Google Docs, and the sheet flips to COVER_LETTER_READY. Below-threshold jobs skip this step automatically.
  5. STEP 5 — Daily digest — both branches merge into one Telegram message: how many new jobs were scanned plus your strong matches with direct apply links.

Status flow in Sheets: NEW → SCORED → COVER_LETTER_READY

Set up steps

You’ll need:

  • n8n (self-hosted or cloud)
  • A RapidAPI key with the JSearch API subscribed — rapidapi.com → search “JSearch” → Subscribe (free tier available)
  • An OpenAI API key — platform.openai.com → API keys
  • Google Sheets + Google Docs OAuth2 credentials — n8n Credentials → Google Sheets OAuth2 / Google Docs OAuth2
  • A Telegram bot — message @BotFather, send /newbot, copy the token; get your chat ID from @userinfobot

Steps:

  1. Import the workflow — in n8n, Workflows → Import from File → select ai-job-hunt-automator.json.
  2. CONFIG — Job Search — open the Set node and fill in:
    • keywords — e.g. AI engineer, Python developer
    • location — e.g. Remote, Bangalore
    • max_results — how many postings to process per run (start with 20)
    • rapidapi_key — paste your RapidAPI key
  3. CONFIG — Google Sheets — create a Google Sheet with a tab named Jobs and headers: Date Found | Job Title | Company | Location | Job URL | Match Score | Verdict | Key Gaps | Status. Paste the spreadsheet ID into spreadsheet_id. Then select your Google Sheets credential in the three Sheets nodes.
  4. CONFIG — AI Scoring — paste your resume text into resume_text (or store it in Drive and reference it), set match_threshold (70 is a good start), and pick your model. Select your OpenAI credential in both chat model nodes.
  5. Cover letter docs — select your Google Docs credential in the Google Docs node; letters are created in your Drive root by default (set a folder in the node if you prefer).
  6. CONFIG — Notifications — paste your Telegram chat ID, select your Telegram credential, and activate the workflow.

Customize it

  • Different job board — swap the JSearch HTTP node for the Adzuna API (free tier) or any jobs API; only the Normalize Job Listings code node needs remapping.
  • Stricter/looser matching — raise match_threshold to 80+ for senior roles, lower it to 60 when exploring new domains.
  • Auto-apply? — deliberately not included: auto-submitting applications can violate job board terms and hurt your reputation. The workflow prepares everything; you click apply.
  • Track applications — add an Applied / Interview / Rejected column to the sheet and update statuses manually for a full pipeline view.
  • Weekly instead of daily — change the Schedule Trigger interval; the dedupe logic keeps it safe at any cadence.
  • Cheaper scoring — gpt-4o-mini is already the default; for very high volumes, batch the scoring step or score only jobs from target companies with an extra IF filter.

Workflow nodes

This workflow is built from 22 nodes. Here's what each one does:

#NodeTypeWhat it does
1 Daily 9AM Trigger Schedule Trigger Starts the workflow automatically on a schedule.
2 CONFIG — Job Search Set Holds editable configuration values — fill in your own settings here before activating.
3 Search Jobs — JSearch API HTTP Request Calls an external API or webhook endpoint.
4 Normalize Job Listings Code Runs custom JavaScript to transform, filter, or compute data between steps.
5 CONFIG — Google Sheets Set Holds editable configuration values — fill in your own settings here before activating.
6 Read Seen Job URLs Google Sheets Reads from or writes to a Google Sheet — this workflow's log and state store.
7 Filter New Jobs Only Code Runs custom JavaScript to transform, filter, or compute data between steps.
8 Log New Jobs — Status NEW Google Sheets Reads from or writes to a Google Sheet — this workflow's log and state store.
9 CONFIG — AI Scoring Set Holds editable configuration values — fill in your own settings here before activating.
10 AI Agent — Score Resume Match AI Agent The AI brain — reasons over the task and calls tools to get it done.
11 OpenAI Chat Model — Scoring OpenAI Chat Model The language model powering the AI agent.
12 Structured Output Parser — Match Score Structured Output Parser Forces the AI to return clean, structured JSON instead of free text.
13 Update Sheet — Status SCORED Google Sheets Reads from or writes to a Google Sheet — this workflow's log and state store.
14 IF — Score ≥ Threshold IF Splits the flow into branches based on a condition.
15 AI Agent — Write Cover Letter AI Agent The AI brain — reasons over the task and calls tools to get it done.
16 OpenAI Chat Model — Cover Letter OpenAI Chat Model The language model powering the AI agent.
17 Google Docs — Save Cover Letter Google Docs Creates or updates a Google Doc.
18 Update Sheet — COVER_LETTER_READY Google Sheets Reads from or writes to a Google Sheet — this workflow's log and state store.
19 Merge — Combine Branches Merge Combines the outputs of two or more branches back into one flow.
20 Build Daily Digest Code Runs custom JavaScript to transform, filter, or compute data between steps.
21 CONFIG — Notifications Set Holds editable configuration values — fill in your own settings here before activating.
22 Telegram — Send Digest Telegram Sends you a Telegram message or alert.

Setup instructions

1. Add your credentials

In n8n go to Credentials → New, create each credential below, then open the named node and select it:

NodeCredential to add
Read Seen Job URLs Google Sheets OAuth2
Log New Jobs — Status NEW Google Sheets OAuth2
OpenAI Chat Model — Scoring OpenAI API
Update Sheet — Status SCORED Google Sheets OAuth2
OpenAI Chat Model — Cover Letter OpenAI API
Google Docs — Save Cover Letter Google Docs OAuth2
Update Sheet — COVER_LETTER_READY Google Sheets OAuth2
Telegram — Send Digest Telegram Bot API

2. Fill in your config values

Open the config node listed below and replace each placeholder with your own value before activating the workflow:

NodeFieldWhat to paste
CONFIG — Job Search rapidapi_key PASTE_YOUR_RAPIDAPI_KEY
Your rapidapi key
CONFIG — Google Sheets spreadsheet_id PASTE_YOUR_SPREADSHEET_ID
Your spreadsheet id
CONFIG — AI Scoring resume_text PASTE_YOUR_RESUME_TEXT_HERE
Your resume text here
CONFIG — Notifications telegram_chat_id PASTE_YOUR_TELEGRAM_CHAT_ID
Your telegram chat id

How this workflow starts

  • Schedule Trigger: Runs every hours.

Use cases

A daily job-hunting assistant that runs every morning while you sleep. The Daily 9AM Trigger fires Search Jobs — JSearch API (RapidAPI, aggregating Google Jobs) with the keywords and location from CONFIG — Job Search. Normalize Job Listings standardizes each posting, Read Seen Job URLs plus Filter New Jobs Only drop anything already processed, and new postings are logged with status NEW. AI Agent — Score Resume Match compares each job against the resume text and returns a structured score from 0 to 100, a verdict from STRONG_MATCH to NO_MATCH, top strengths, and key gaps via Structured Output Parser — Match Score. The sheet flips to SCORED; jobs at or above the threshold pass IF — Score ≥ Threshold (default 70), get a tailored cover letter of max 250 words referencing real achievements from AI Agent — Write Cover Letter, saved to Google Docs, and the sheet flips to COVER_LETTER_READY. Both branches merge into Build Daily Digest, and Telegram — Send Digest delivers the count of new jobs plus strong matches with direct apply links. It deliberately stops before applying — auto-submitting can violate job board terms, so the workflow prepares everything and you click apply.

The laid-off software engineer running a full-time search

An engineer laid off in a restructuring is treating the search as a job, but scrolling job boards for three hours a day is not a system. With max_results at 20 and keywords like "backend engineer," the workflow scans, dedupes, and scores every morning; the Telegram digest tells him there were 14 new postings, 3 strong matches, and here are the apply links with cover letters already sitting in Google Docs. The Key Gaps column doubles as interview prep — he can see exactly which required skills he lacks before the recruiter call. Practical tip: raise match_threshold to 80 once the pipeline is full, so the digest only surfaces roles worth the application effort.

The career switcher testing new waters

A support engineer wants to move into AI engineering but does not know which postings are realistic. The verdict scale (STRONG_MATCH down to NO_MATCH) plus the strengths and gaps output turns each morning's digest into market feedback: after two weeks she can see which of her skills transfer and which gaps repeat. The resume text in CONFIG — AI Scoring is the lever — she rewrites it to emphasize projects, and the scores move. Practical tip: lower the threshold to 60 while exploring. At the exploration stage, borderline matches are information, not noise.

The final-year student running a structured hunt

A student applying to forty companies keeps everything in memory and loses track of who replied. The sheet's status flow — NEW to SCORED to COVER_LETTER_READY — is already a pipeline; adding Applied, Interview, and Rejected columns (per the customize guide) turns it into a full tracker with zero extra tooling. The dedupe logic (Read Seen Job URLs plus Filter New Jobs Only) means reposted jobs never clutter the list, which matters when campus season reposts the same roles weekly. Practical tip: if JSearch coverage is thin for her market, swap the search node for the Adzuna API — only the Normalize Job Listings code node needs remapping.

The employed job seeker with fifteen minutes a day

A working professional cannot job-hunt at a desk all day; she has the train ride. The digest design fits that constraint exactly: one Telegram message with the day's strong matches and apply links, cover letters already written and saved. She reads the digest, opens the top two matches, clicks apply, done. Because the workflow never auto-applies, there is no risk of an accidental application to her current employer's posting — she sees everything first. Practical tip: change the Schedule Trigger to weekly if the daily cadence produces more digest than she can act on; the dedupe logic keeps any cadence safe.

The remote-only job seeker filtering out the noise

A designer who will only work remotely was drowning in "hybrid, three days in office" postings. Setting location to Remote in CONFIG — Job Search filters at the source, and the AI scoring then ranks what remains against the resume rather than the commute. The digest becomes a short list of genuinely applicable roles instead of a long list of maybes. Practical tip: for very high volumes, add an IF filter before scoring that only scores jobs from target companies — scoring is the expensive step, and gpt-4o-mini is cheap but not free.

← All templates