New The Skills of Tomorrow: how AI-exposed is every skill in 2026? See the data →
Hiring intelligence as a JSON feed

28,129 enriched job postings, 932 companies, 30+ public endpoints.

Every chart on yubhub.co/live and every section on company intel pages is backed by a public REST endpoint. Built for talent intelligence, equity research, and BAs who treat hiring as a leading indicator. Talk to us if your team needs the full corpus, higher rate limits, or a custom feed.

Talk to us openapi.yaml ↗
28,129
Indexed roles
932
Companies
30+
JSON endpoints
CC BY 4.0
Open data licence
The Data
What's actually in the corpus
Every posting is sourced from a verified employer careers page (no aggregator scrape, no LinkedIn). Discovery first attempts the ATS API (Greenhouse, Lever, Workable, Pinpoint, Oracle HCM) for full content, falls back to sitemap parsing, then Firecrawl. Every successful posting then runs through Llama 3.1 8B for structured extraction.
FieldTypeNotes
titlestringCleaned, original-case
company + company_*linkedWikidata + FMP + GitHub + 7 other free sources, daily-refreshed
locationstringFree text; resolved to one of 12 region buckets in API responses
work_arrangementenumremote / hybrid / on-site / unknown
experience_levelenumentry / mid / senior / staff / executive / intern
salary_min / salary_max / salary_currencystructuredUSD, GBP, EUR, AED — not all roles disclose; use salary-transparency endpoint to see disclosure rate per industry
skills.required / skills.preferredarraysCanonicalised via 240-row alias map (k8sKubernetes, llm/llmsLarge Language Models, cfdComputational Fluid Dynamics, etc.)
category / industrystringsAI-tagged; collapsed into 12-bucket canonical map for cross-company comparison
soc_code + ai_exposurenumericO*NET 30.2 SOC crosswalk × Anthropic Economic Index 2026-03-24 — a 0–1 score per role
discovered_at + enriched_at + expired_attimestamps (ms)Full lifecycle. Time-to-fill is computable as expired_at − discovered_at
Endpoints in use
Everything on YubHub is API-first
The page you're reading is the only HTML surface that doesn't fetch from these endpoints. Each route below is what powers a section you can see in production.
SurfaceEndpointReturns
/live hero/stats/overviewTotal jobs, companies, feeds, data range
/live momentum sparklines/stats/industry-momentumPer-industry accelerating / stable / decelerating signal
/live Skills of Tomorrow/stats/skill-ai-exposurePer-skill AI-exposure score, bucketed amplified / native / threatened / shielded
/live Money chart/stats/salary-by-sectorMin / median / max per category, with sample sizes
/job-statistics/stats/market-pulse + 16 othersCombined market-health indicators, all charts
/companies/anthropic → 11 sections/stats/company/:slugProfile + velocity, momentum, salary vs industry, skills LQ, geographic shift, executive hires, AI exposure, peer set, posting dynamics, market share
26-week trajectory chart/stats/company/:slug/timeseriesPer-company time series, dispatched by metric
Peer-set table/stats/company/:slug/peersSame-industry peers ranked by Jaccard similarity, each annotated with momentum
Future /companies hub/stats/companies/leaderboardCross-company rankings — accelerating / AI-exposed / hiring-velocity / exec-hires
Hiring trends pages/stats/by-title/:titlePer-title deep stats: top companies, skills, salary bands, locations

Full machine-readable specification: openapi.yaml (OpenAPI 3.0).

Use Cases
What teams are using this for
Concrete examples drawn from analyst conversations. Some require the API tier (rate limits, full corpus); some work fine on the public endpoints.

Talent Intelligence

Build watchlists of companies in your competitive set. Track who's hiring up, who's hiring down, where executive-level hires are clustering. The 14-day momentum signal flags shifts before they show up in headlines.

leaderboard?metric=accelerating · /company/:slug/peers

Equity Research

Hiring is a leading indicator that lags earnings reports by months. New emerging-hub flags ("AWS just opened 12 EMEA postings, up from 0") often precede strategic announcements. Combine with seniority_anomalies to spot leadership-tier buildup.

/company/:slug#geographic_shift · seniority_anomalies

VC + Market Mapping

Industry-momentum sparklines surface which sectors are heating up week-over-week. Skills LQ shows what disproportionate capabilities a portfolio company is building vs its peer baseline — useful for diligence and post-investment monitoring.

/stats/industry-momentum · /company/:slug#skills_lq

HR Vendor Enrichment

If you're shipping HR or talent-stack tooling, you can enrich your customer's view with our skills, salary bands, AI-exposure scores, and geographic distributions. Free for the public corpus; API tier for the full 12k-row firehose.

/stats/skill-ai-exposure · /stats/salary-by-sector

Access
How you get the data
Three paths. Most analyst use cases sit comfortably in the free tier; the API subscription unlocks higher rate limits and authenticated endpoints; custom feeds are bespoke.

Open & free

$0
  • Every /stats/* endpoint
  • OpenAPI spec + JSON-LD Datasets
  • Markdown alternates (Accept: text/markdown)
  • Public rate limit (Cloudflare-edge-cached)
openapi.yaml ↗

Custom feed

Bespoke pricing
  • YubHub-managed scrape of your own careers / aggregator surfaces
  • JBoard-compatible XML or jobs.json output
  • SLA + dedicated support
  • Already in production with publishers and recruiters
See feed plans →
Sample
A request that returns real data
No keys needed for the public stats family. Try this and you'll have the complete shape of a company intelligence payload.
# Single-company intelligence — profile + 11 benchmarking sections
curl 'https://api.yubhub.co/stats/company/anthropic' | jq

# 26-week posting volume time series
curl 'https://api.yubhub.co/stats/company/anthropic/timeseries?metric=velocity&weeks=26' | jq

# Cross-company leaderboard — fastest-accelerating in last 14 days
curl 'https://api.yubhub.co/stats/companies/leaderboard?metric=accelerating&limit=10' | jq

# Markdown brief — content-negotiated for agent ingestion
curl -H 'Accept: text/markdown' 'https://yubhub.co/companies/anthropic'
Methodology & Limits
What you should know before quoting numbers

Coverage. Roles are sourced from individual employer careers pages, currently UK and global. Discovery is best-effort; we miss roles that aren't on a careers site. Total corpus is in the hero stat band.

Refresh cadence. Free / Basic feeds refresh weekly; Publisher / API tiers refresh daily. Company-level intel (profile, news, GitHub, SEC) re-enriches nightly.

Salary parsing. Where the source posting includes a structured range we extract salary_min, salary_max, salary_currency, salary_period. Disclosed ranges are a minority — see /stats/salary-transparency for per-category disclosure rates. Don't quote a "median salary for X" without the disclosure caveat.

AI exposure. Every posting is mapped to a SOC code via O*NET 30.2; the SOC code carries an exposure score from the Anthropic Economic Index 2026-03-24 release. The number is a probability, not a verdict — read it as "this kind of role is exposed to current LLM capability," not "this role is being automated."

Edge caching. Most endpoints have a 5-minute edge cache. Hit them again with a unique query param if you need fresh after a deploy. Custom rate limits are a paid feature.

Get In Touch
Tell us what you're trying to build
Most useful conversations start with "we're trying to answer X about Y companies." We'll come back with whether the public endpoints handle it, whether you need API-tier access, or whether a custom feed makes sense.
Talk to us Explore Company Intel →