JobMatch AI - A Slack bot that finds jobs and writes the resume

Tailoring a résumé for every job is exhausting

Every serious application needs a custom CV, but doing that for thirty roles a week, by hand, is impossible.

Five simple steps

No technical knowledge needed to follow this, every box is a thing happening in the background, automatically.

Watch it run

A short clip showing one Slack message turn into three tailored applications.

The full workflow

Here's the entire automation laid out end-to-end. The three stages below break this same canvas into bite-sized pieces.

Complete n8n workflow - Slack trigger to job-channel reply

~20 nodes · 3 stages

The whole pipeline in one canvas. Read left-to-right: a Slack mention triggers it, and every job from the search loops through tailoring and delivery.

1

Listen & Search

Slack trigger, query validation, JSearch API call, result check, split into single jobs.

2

Tailor

Read master résumé, AI rewrites bullets, structured output, duplicate check, rate-limit wait.

3

Deliver

Copy template, inject content, optional Gmail draft, post summary back to Slack.

Three stages working together

Same canvas as above, broken into the three things it actually does.

Stage 1 - Listen & Search

filters bad queries

• 1

  Listens for the mention

  Slack triggers the workflow when someone tags the bot with a query.

• 2

  Validates with Gemini

  An AI classifier decides if the message is a real job query or noise. Junk stops here.

• 3

  Sends a "Searching…" prompt

  Posts back to Slack so the user knows the bot is working - not silently stuck.

• 4

  Hits the JSearch API

  Fetches fresh job listings, typically three to five per query.

• 5

  Checks the results

  If the API returned nothing useful, the bot says so instead of pretending it found jobs.

• 6

  Splits into one-at-a-time

  Hands the jobs off one by one so the next stage can process each cleanly.

Stage 2 - Tailor

résumé per job

• 1

  Reads the master résumé

  Pulls the latest content from a Google Doc - the single source of truth for the CV.

• 2

  AI rewrites the bullets

  An OpenRouter agent reads the job description and tailors bullets and the proposal to match.

• 3

  Structured output, every time

  A schema parser keeps the shape consistent. No prompt drift, no broken docs.

• 4

  Checks Drive folder

  Confirms the destination folder exists before trying to save anything to it.

• 5

  Avoids overwriting

  If a résumé for this exact role already exists, it skips and moves on.

• 6

  Pauses between calls

  A small wait stops the workflow tripping over Google Drive's rate limits.

Stage 3 - Deliver

email + Slack

• 1

  Copies the template

  A fresh Google Doc is created and named for the role. Formatting stays consistent across every output.

• 2

  Updates the document

  Injects the AI-tailored bullets and proposal into the new copy.

• 3

  Decides on the email

  Checks whether the listing actually has a recruiter email to write to.

• 4

  Drafts a Gmail (if applicable)

  If there's a recipient, a draft is queued - ready to review and send.

• 5

  Posts back to Slack

  One clean message per job: title, apply link, résumé link. All in the channel.

• 6

  Loops to the next job

  Repeats until every job from the original search has its own résumé and Slack post.

One message in, four artefacts out

No custom UI to learn - just type a search. Everything that comes back is something you can act on immediately.

Things you can ask

Job searches

• Data entry jobs
• Remote VA roles

Specific titles

• Automation Engineer NZ
• n8n developer remote

Or anything in between

• Part-time bookkeeping
• Customer support remote NZ

Trigger

Slack @mention

AI Models

Gemini · OpenRouter

Job Source

JSearch API

Output

Doc + Email + Slack

What changed for the workflow

Tools used

Familiar, reliable platforms, no exotic tech.