Cleanjobdata

CleanJobData (operated by Jobr LLC) — Normalised global job-postings corpus (3.3M+ indexed listings) pulled directly from ATS platforms and employer career sites rather than job boards, delivered as one stable JSON schema: per-posting title, employer identity (stable employer_id with logo, website, company LinkedIn, industry, headcount, founded year, socials, and named team members with LinkedIn profiles), structured location (country/state/city IDs plus lat/lng, country code and timezone where resolvable), salary parsed into min/max/currency, remote flag and remote_type, employment_type, exp…

Cleanjobdata offers (Alternative, Sentiment) — Published: 3.3M+ job listings indexed in total, 36,280 new jobs posted per day (7-day average), with live 'Added Past 30 Days' and 'Total Jobs Indexed' counters on the homepage; the US feed (country_id=233) is described as the largest-volume region in the index, and coverage is global (sample records include Spanish employers such as Mercadona). With ~15+ structured fields per record including full description text, the licensed bulk export is a multi-GB-scale daily-growing corpus..

LLM pre-training and fine-tuning corpora on clean structured job text with normalised schema across many ATS layouts (explicitly a licensed purpose); job-title/salary/seniority extraction model training and evals; wage and compensation benchmarking by title, seniority and metro from parsed salary bounds; labour-demand nowcasting and hiring-signal alpha by employer, sector and metro (per-employer daily posting velocity); account and sales intelligence (which employers are hiring which skills, where teams are expanding); job-board and HR-product content feeds; workforce-planning and skills-demand analytics; recruiter matching and talent-supply models using the resolved employer graph plus named team/hiring-contact data.

The data is with 2 years of history.

Coverage spans US, UK, Europe, Other; Human Resources & Employment Services; alternative, sentiment; equities.