From massive crawl volumes to the strong matches we surface — the full pipeline, in real numbers.
From everything we crawl, to the few that match. Here's how the pipeline narrows.
Each source is a separate integration. Some are aggregators spanning multiple regions, some are niche specialists.
Six steps from crawl to match.
n8n orchestrates per-source crawlers, hitting each on a configurable interval. Some return 100 jobs per page, some hit single-API limits.
Same job on multiple boards collapses to one. URL match catches exact reposts, title+company catches recruiter variants.
Each search profile applies its own filters: title patterns, exclusions, country gate, source toggles, minimum description length.
Jobs that pass all filters land in the pipeline. Everything else stays in sandbox for 48h, visible in the admin for debugging.
Each job is evaluated against the user's career profile. Returns a 0–100 fit score plus structured analysis of matches and gaps.
Jobs scoring ≥80 are surfaced as strong matches. Below 80 stay accessible but aren't pushed forward as primary recommendations.
Real job market data from sources across Europe and the US — what postings actually show, not what surveys estimate.
Data collection started April 16, 2026. Numbers grow daily as more jobs are analyzed. Skill demand, country deep dives, and trend analysis coming in future Market Pulse editions.
Percentage of job postings that include salary information. The EU Pay Transparency Directive requires salary disclosure from June 7, 2026.
Based on live job postings across {sources_count} sources. Single-platform reports show higher rates — this is what job seekers actually encounter across the broader market.
How each country's job market compares to the EU average. Values show over- or under-representation vs the cross-country mean.
Every match goes through this AI pipeline. Here's the scale.