LinkedIn Job
Dataset

Plus job_applicant_history as its own table — one row per observed change in a posting's applicant count, joinable on linkedin_job_id.

psql "postgresql://ds_a41f9c:••••@db.linkfetch.io:5432/linkfetch_public?sslmode=require"

linkfetch_public=> \dt
 Schema |         Name          | Type
--------+-----------------------+-------
 public | jobs                  | table
 public | job_applicant_history | table
 public | dataset_meta          | table

-- Who started hiring engineers this month, and where?
SELECT company, region, count(*) AS openings
FROM jobs
WHERE first_seen_at > now() - interval '30 days'
  AND job_function ILIKE '%Engineering%'
  AND is_active
GROUP BY company, region
HAVING count(*) >= 5
ORDER BY openings DESC;

Indexed for the obvious things — company, title, location, recency, salary floor, and full-text over descriptions — so ordinary queries stay fast without you tuning anything. Your role is read-only, with a per-session query timeout so a runaway aggregation only ever costs you your own connection.

We are not going to out-scale Coresignal on volume or Bright Data on history, and the last two rows say so. What we have is the shape of the deal: no meter, no minimum, no integration work, and two columns of hiring signal they don't carry. Competitor details are from their public pricing as of August 2026 and change without notice — check them.

Need a filtered cut, an invoice, or a DPA? Talk to us. Prefer per-lookup API calls over bulk? That's the API.

• What exactly do I get for $199?

  The entire database as files, built fresh the day you buy: a PostgreSQL custom-format dump (pg_restore and go), plus CSV and JSONL of the same rows for everything that isn't Postgres. Both tables ship — the jobs archive and the applicant-count time series — with a manifest stating exact row counts and the coverage window. It's a one-time purchase; the files are yours permanently.

• And the $49/month live access?

  A read-only PostgreSQL connection string to a database holding the same data, rebuilt from the live scrape every night. Connect with psql, DBeaver, your ORM or a notebook, and query as much as you like — no credits, no per-row charges, no meter. It's the plan B for when a frozen dump isn't enough: most buyers start with the dump.

• What is the applicant-count time series, and why does it matter?

  Every posting is re-visited continuously, and each time the applicant count changes we record the new value with a timestamp. That turns a static job board dump into lifecycle data: how fast a role accumulates applicants, which roles stall, and — combined with first_seen_at and is_active — roughly how long a posting stays open. A single-snapshot dataset cannot answer any of those questions, and as far as we know no competitor sells the series.

• How much history is there? Be honest.

  The archive starts in late April 2026, so this is months of depth, not years. Coresignal and Bright Data have been collecting far longer and will beat us on historical reach. What we have that they don't is applicant counts, the applicant time series, flat pricing and direct SQL access. If your work needs a decade of postings, buy from them; if it needs live signal you can query without a meter running, buy from us.

• Is there personal data in here?

  No, and there deliberately never will be. This is job and company data only. We store no LinkedIn profiles, run no fake accounts, and access nothing that isn't publicly visible. Job descriptions sometimes contain a recruiter's email or phone number, so every description is passed through a redaction step before it reaches the public database or a snapshot, and the export aborts if anything person-shaped survives the check.

• Can I resell it?

  No. Both SKUs are licensed for your own use and analysis, including commercial use, but redistributing or reselling the dataset in whole or in substantial part is not permitted. Building a product on top of it that exposes derived results is fine. Republishing the rows themselves is not.

• How fast is delivery?

  You get a confirmation email the moment payment clears, then a second email with the goods. Snapshots are built fresh from the current database and the download links are emailed automatically — usually within the hour, always within 24. Live access works the same way: a dedicated read-only role is provisioned and the connection string emailed. A human watches every order and steps in if anything fails.

• Can I try before buying?

  Yes. There's a free 1,000-row sample below with the complete schema and full description bodies. No signup, no email gate — it's a static file. It is drawn at random rather than hand-picked, so its field coverage matches the real dataset.