See llms.txt for all machine-readable content.

Back to Templates

Find course gaps from LinkedIn jobs with OpenAI, Apify and Slack

Quick overview

Find the skills employers keep asking for that the LinkedIn Learning course catalog barely covers. Each month it reads LinkedIn postings for your roles, has AI name the skills, counts them in the text, checks the catalog, and drafts a course idea per gap for Google Sheets and Slack.

How it works

  1. A monthly schedule starts the run, and a Set node holds your roles, the place and the gap rules.
  2. The first Actor searches LinkedIn jobs for each role and returns LinkedIn's own job count for the month.
  3. The same Actor reads every posting in full. Reposts with identical text are kept once.
  4. An Information Extractor names the 20 most common skills and the exact words the postings use for each.
  5. A Code node counts how many postings mention each skill, overall and per role. Skills above your cut-off in any role go to the catalog check.
  6. The second Actor searches the LinkedIn Learning catalog for every in-demand skill in one run.
  7. A Code node counts the courses whose title names the skill and finds the newest one. Few courses, or only old ones, is a course gap.
  8. A second Information Extractor drafts a course idea for each gap from lines in the postings. Every skill goes to Google Sheets, and the top gaps go to Slack.

Setup

  1. Create a free Apify account, copy the API token from Settings, Integrations, and add it as an Apify credential on the three Actor nodes.
  2. Add an OpenAI credential on the two chat model nodes, or swap in any other chat model.
  3. Connect Google Sheets and pick a spreadsheet and tab. The columns are created from the first rows.
  4. Connect Slack and pick a channel for the digest.
  5. Put your roles and place in the Set node, run it once by hand, then activate it.

Requirements

  • An Apify account; a month of three roles costs about one cent in Actor charges
  • An OpenAI API key, or any chat model credential n8n supports
  • Google Sheets and Slack credentials
  • No LinkedIn account, LinkedIn Learning license or scraping key

Customization

  • roles takes any comma-separated list of job titles; one search runs per role
  • postingsPerRole sets how many postings are read per role; more give steadier shares
  • minShare is the share of postings a skill needs in at least one role to be checked
  • thinCourses and staleYears decide what counts as a gap: too few courses, or none recent enough
  • ideasInSlack caps how many course ideas the digest shows
  • Swap Slack for the optional Gmail node, or add the optional details read for the closest course's ratings and syllabus

Additional info

How do I find course ideas that people will pay for? Start from what employers ask for, then check what already exists. This template reads recent LinkedIn postings for the roles your audience wants, counts the skills those postings mention, and searches the LinkedIn Learning course catalog for each one. A skill that many postings want and few courses teach is a course gap. That's the topic to build first.

How is a gap decided? A course counts only when its title names the skill, because the catalog search is loose: "go-to-market strategy" reports hundreds of results, but in our test only one course had it in its title. A skill with two or fewer such courses, or none released in the last three years, is a gap. Both limits sit in the Set node. When no title names the skill, the row shows the closest course so you can judge the near miss yourself.

Can I trust the numbers? The model only names the skills and drafts the course ideas. Code does the counting, so "8 of 20 postings" means 8 posting texts contain the words, and "1 course names it" is a count of real catalog rows. A skill the model names but no posting uses is marked Not in the postings and never reaches the catalog check. An idea for a skill that is not on the gap list is dropped.

Why is one employer not counted three times? Companies often post the same job in several cities with identical text. The workflow keeps one copy of each description before counting, so a single employer's reposts don't make a skill look popular.

Who is it for? Course creators and instructors choosing their next topic, bootcamps checking a syllabus against the market, L&D teams deciding what to buy or build, and training companies that sell to one industry. Put the roles your learners want in the Set node.

What does it cost? Every posting, listing and course costs $0.00001, so three roles at ten postings each, plus twelve catalog searches, come to about one cent a month in Apify charges. The two model calls cost a few cents on gpt-5-mini. A run takes a few minutes.

What else can these Actors do? LinkedIn Job Search Scraper is a LinkedIn job scraper and job search API in one: keyword and place search with date and company filters, full job descriptions, result counts, and title, company and location autocomplete, all without a LinkedIn login. LinkedIn Learning API searches the LinkedIn Learning course catalog by keyword, topic, level, length and software, and returns full course records with ratings, reviews, syllabus and learning paths.

About Apify. LinkedIn Learning API runs on Apify, a platform of ready-made scrapers and APIs for sources that have no usable public API. Actors are pay per use with no subscription, and they run in the cloud, so nothing runs on your machine and nothing breaks when the source changes its markup.

New to Apify? You can set up a free account and get the API token this template needs in a couple of minutes.