Know when a European company starts hiring, wins a tender or posts pay.
Every day we read public information about European companies (who they hire, which public contracts they win, what pay they post, what tools they use) and deliver what changed to the people who need it.
REST APIMCP for AI agentsDaily filesNo personal data
openings_7deffektiv.se+14 in 7d · 53 open · jobtechSE
posted_paytrmlabs.com€94k–€161k a year · Business & adminGB
tender_awardedTeleperformance Netherlands B.V.€31.1M · Business servicesNL
posted_payfonio.ai€65k–€100k a year · Business & adminES
Real rows from the free samples. Awards from EU TED and UK notices; pay, openings and tools from employer job boards; spikes from the hotel rate curve.
Questions it answers
Which German companies started hiring in a new country this week?GET /v1/pulse/events?country=DE&types=new_country
Who won French public contracts above €1M this month, and are they hiring?GET /v1/tenders?country=FR&kind=award&minValueEur=1000000
Which of my 500 target accounts posted a salary range for the first time?GET /v1/watchlists/{id}/changes
Four ways to use Arcadium Data.
The same data, bought the way that fits. Most customers start with change signals for a list of companies.
Monthly hiring history of listed companiesback to 2023, coming
How it works.
01
We read public sources every day
Employer career sites and applicant tracking systems, official tender portals (EU TED, UK Find a Tender, French, German, Polish and Bulgarian boards), Shopify storefronts and public review pages. No logins, no personal profiles.
02
We join them on the company domain
A job board, a contract winner and a review page all resolve to one website domain. Each change becomes a dated event: what happened, when it happened, when we first saw it, and the evidence.
03
You query, track or download
Ask the REST API for one company or a filtered list, put up to thousands of domains on a watchlist and read only what changed, let an AI agent call the MCP server, or take the full daily files.
The changes we detect: thirteen event types, one key.
Each event carries the day it happened and the moment we first saw it, so a backtest only sees what was knowable then. provisional while the baseline is under 28 days.
hiring_upemployer job boardssince 21 Oct 2026
hiring_downemployer job boardssince 21 Oct 2026
new_countryemployer job boardssince 21 Oct 2026
new_functionemployer job boardssince 21 Oct 2026
ats_switchemployer job boardssince 21 Oct 2026
pay_first_seenposted pay in job adssince 21 Oct 2026
tech_addedtools named in job adssince 21 Oct 2026
tech_removedtools named in job adssince 21 Oct 2026
shop_app_installedShopify storefront scanssince 21 Oct 2026
shop_app_removedShopify storefront scanssince 21 Oct 2026
review_velocityreview panelsince 21 Oct 2026
rating_changereview panelsince 21 Oct 2026
tender_awardedEU TED and UK award noticessince 23 Oct 2025
Point Claude, Cursor or any MCP client at the server. Every answer has an as_of date, a source and coverage notes. $0.01 a call, no seat; free during the preview, 25 calls a day.
API plans are for querying what you need: companies, searches and the changes on your watchlist. They do not hand over the whole database; that is what the daily feeds are for.
API Free
€0
1,000 API calls a month
100 tracked companies: their changes arrive on your watchlist
Every plan queries the same 8 datasets: Company Pulse events, company profiles and hiring, job search, public tenders and awards, posted-pay bands and disclosure, Shopify store apps, review velocity, hotel rate curve. One request is one call and returns up to the plan's row limit; the remaining quota is in the x-quota-remaining header. At the limit the API stops until the 1st; there is no overage bill.
Want the whole dataset every day? That is a feed, priced separately.
EU TED, UK Find a Tender and national tender portals; employer career sites and ATS boards; public review and store pages. Every row names its source.
No personal data
Companies only. No names, emails or phone numbers of people; job text is filtered before it is stored. There is nothing to opt out of.
Point in time
Every event carries the day it happened and the moment we saw it, so a backtest only uses what was knowable then.
EU-based
Run from the EU, priced in euros, invoiced with VAT where it applies.
Questions
Do I get the whole database?
Not with an API plan. An API plan lets you query what you need: one company, a filtered search or the changes on your watchlist, up to your monthly calls. If you want everything, every day, take a data feed: the full daily file of a dataset (jobs, tenders, pay), priced separately and delivered to your bucket.
What does a €49 Starter plan buy?
25,000 API calls a month across every dataset, 1,000 companies on watchlists (their changes arrive without a lookup), 60 requests a minute and up to 100 rows per call. It does not include the bulk files.
How fresh is it?
Sources are read every day: employer job boards daily, tender portals each morning, Shopify stores on a weekly rotation. Every answer carries an as_of date, so you always know how old a fact is.
How far back does it go?
Daily company events since October 2026, public tenders and awards since October 2025. Monthly hiring history back to 2023 for listed companies is being built and is not sold yet.
Where does the data come from?
Public sources read without logging in: official open-data tender portals (EU TED, UK Find a Tender, France BOAMP, Germany, Poland, Bulgaria), employer career sites and applicant tracking systems, Shopify storefronts and public review pages.
Is there personal data?
No. We store companies, jobs, contracts and aggregates, never names, emails or phone numbers of people. Job descriptions are filtered for contact details before they are stored.
Can my AI agent use it?
Yes. The MCP server works with Claude, Cursor and any MCP client; every tool returns compact JSON with an as_of date, the source and coverage notes. There is also an OpenAPI 3.1 file and llms.txt.
Monthly reports, from the same data.
Free, citable numbers with the CSV and the method.