Hiring-intent signals read from company ATSes and public job boards across 7 regions. Median 2.1 hours from publish to detection on ATS-direct sources (Lever 12 min, Greenhouse 1.7 h); 9 hours across the whole corpus (measured 2026-08-08). How this is measured.
Category listings are the sum of the per-category counts in the directory below. The categories overlap — a req that matches a role category and a sector category is counted in both — so this is more than the number of distinct reqs behind it.
Every page here is rebuilt every 3 hours from the live corpus, so the counts move with the market rather than with a monthly export.
590 tracked markets · 455,507 category listings in the last 30 days · 7 regions of coverage.
The actual posting and the board it was detected on — pulled straight from the live corpus, never inferred intent.
Call one MCP tool and get back the reqs your ICP has opened most recently.
Every result is a real posting we observed, tagged with the board it was detected on — never inferred intent, never a link to a competitor's listing.
curl -X POST https://mcp.reqbeat.com/mcp \
-H 'Content-Type: application/json' \
-d '{"jsonrpc":"2.0","id":1,"method":"tools/call",
"params":{"name":"who_is_hiring_for",
"arguments":{"role":"backend engineer","geo":"USA","limit":5}}}'
import requests
resp = requests.post(
"https://mcp.reqbeat.com/mcp",
json={
"jsonrpc": "2.0", "id": 1, "method": "tools/call",
"params": {
"name": "who_is_hiring_for",
"arguments": {"role": "backend engineer", "geo": "USA", "limit": 5},
},
},
)
print(resp.json())
{
"mcpServers": {
"reqbeat": { "url": "https://mcp.reqbeat.com/mcp" }
}
}
Paste into any MCP client, then ask it who is hiring for your ICP — no key needed for this tool.
Start free — no credit card. Pay only for the signals you pull; scale up when you're ready.