Live · updated every hour

The data, exactly as it is

What we have, how much of it is filled in, how fresh it is, and how it was collected. If a number here looks small, it is because it is true. Every figure on this page comes from the same tables the API reads.

Ady, the crawler, is runningprofiling around the clock
crawled chelsearow.com just nowcrawled chelseapost.com just nowcrawled chelseapsychotherapynyc.com just nowcrawled chelseapratchlerphotography.com just nowcrawled chelseacontreras.com just nowcrawled chelseaconstable.com just nowcrawled chelseapremium.com just nowcrawled chelseascreativeconfections.com just nowcrawled chelsearosephotog.com just nowcrawled chelseamills.com just nowcrawled chelseasculinaryindulgence.com just nowcrawled chelseaconstructioncorp.com just nowcrawled chelseaconaboy.com just nowcrawled chelsearobinsonrealestate.com just nowcrawled chelseaconradt.com just nowcrawled chelseascreename.com just nowcrawled chelseaprivateoffice.com just nowcrawled chelseaprivateequity.com just nowcrawled chelseascreativeconcepts.com just nowcrawled chelseamemmolophotography.com just nowcrawled chelsearow.com just nowcrawled chelseapost.com just nowcrawled chelseapsychotherapynyc.com just nowcrawled chelseapratchlerphotography.com just nowcrawled chelseacontreras.com just nowcrawled chelseaconstable.com just nowcrawled chelseapremium.com just nowcrawled chelseascreativeconfections.com just nowcrawled chelsearosephotog.com just nowcrawled chelseamills.com just nowcrawled chelseasculinaryindulgence.com just nowcrawled chelseaconstructioncorp.com just nowcrawled chelseaconaboy.com just nowcrawled chelsearobinsonrealestate.com just nowcrawled chelseaconradt.com just nowcrawled chelseascreename.com just nowcrawled chelseaprivateoffice.com just nowcrawled chelseaprivateequity.com just nowcrawled chelseascreativeconcepts.com just nowcrawled chelseamemmolophotography.com just now

What we have

Each tile says how to query it. Company, tech and signal data needs no key.
132Mcompanies indexedEvery root domain we know exists. 23M carry a full profile; the rest are profiled on request in about two minutes.GET /api/v1/lookup?domain=…23Mcompanies profiledName, tech stack, email pattern, signals and, where found, contacts. This is the set search, /companies and the list endpoints run over.GET /api/v1/companies?sector=…126.6Mtechnology detections · 4,077 technologiesDetected from live page source, headers and DNS with Wappalyzer signatures. Never self-reported.GET /api/v1/tech/intercom15Memail addresses foundEvery address carries a verification status; the verified count appears here after tonight's snapshot.GET /api/v1/lookup?domain=… (free key)5.3Mpeople foundNames and roles from team, about and author pages, with a confidence score. We show people at confidence 40 and above.GET /api/v1/people?domain=… (free key)16Mlocal businessesBy city and category: restaurants in London, dentists in Austin. Address, phone where public, website, hours.agentdata.run/local-businesses

How much of a profile is filled in

Share of profiled companies with each field present. Measured nightly, 11 Oct.
Company name
…
At least one technology
…
Industry classified
…
Email pattern
…
At least one email
…
A verified email
…
At least one person
…
Country known
…
Amber means we know it is thin; red means we are not close yet and say so rather than guess.

Email verification, honestly

15,815,248 addresses found. The verified, queued and dead split appears after tonight's snapshot.Verification runs continuously and every address carries its status and the date it was checked, so you can filter on it. We do not call an address verified unless SMTP said so.

Freshness

Age of each profile's last crawl.
…
under 7 days
…
7 to 30 days
…
30 to 90 days
…
older
Anything you ask for that we have never seen is profiled in about two minutes; a profile older than 30 days that someone opens is re-queued.

Profiles per day, last 14 days

crawl_yield_daily
Daily yield appears here once the crawl queue has a day of history.Profiled means the crawl returned a usable company record. The outcome log started on 16 September, when the crawler moved to a scored queue; days before that are empty here, not zero. The chart will say so when it is not flattering, too.

Where the companies are

Top industries among classified profiles.
E-commerce & Retail
599K
Professional Services
579K
Media & Content
443K
Hospitality & Travel
418K
Healthcare & Life Sciences
394K
Education & Learning
179K
Industry is set on a fraction of 23M profiles so far, classified from the site's own description. Country is known for far fewer; we say so rather than guess from the domain ending.

What we do not have

✕ Mobile numbers. We do not collect or buy them.✕ Revenue, funding rounds or headcount from filings. Size is estimated from the website.✕ LinkedIn profile scraping. A person's LinkedIn URL is stored only when the company's own site links to it.△ Country and city for most companies. Being filled in from DNS, contact pages and local business listings.△ Verification for most email addresses. The checker runs around the clock; the split above is the honest state.

How it is collected

✓ Public web pages only: the company's own site, DNS and MX records, HTTP headers.✓ robots.txt respected on every visit; one crawler identity, agentdata.run/bot.✓ Emails are found on pages, never purchased. Role addresses (info@, support@) are shown to everyone; personal ones need a free account.✓ Every record shows where and when it was found. Removal on request: data removal.

What a company record contains

Field by field, with where each one comes from.
GET /api/v1/lookup
Companyname, domain, sector and vertical (LLM-classified from the site's own description), business model, B2B or B2C, size band, country where known
Tech stackevery detected tool with category, detection method and evidence; first and last seen
Email patternthe address format the company uses (first.last, flast, first), with confidence
Emailseach address with confidence, verification_status, verified_at, where it was found and when, role-based flag
Peoplename, title, seniority, department, decision-maker flag, confidence; LinkedIn only if the company links it
Signalspricing page, free trial, API docs, blog, careers page, primary call to action
Changestools added, removed or switched, with the date we saw it
Provenancelast crawled, profile completeness score, crawler outcome
Nothing is inferred and presented as fact: a field is null when we did not find it. The full response shape is in the API docs.

What is next

In the order we are doing it.
01  Clear the email verification backlog, so the verified share climbs and the split above shows it.02  Country and city for every profiled company, from contact pages, DNS and local listings, so the red bar above stops being red.03  Search by company name, not just domain: type “Stripe” and land on stripe.com.04  Weekly public changelog of the numbers on this page, so you can see them move.
Something you need that is not here? Tell us; the queue is short and we answer.

Query any of it

No account for company, tech and signal data (300 requests per 5 hours per IP). A free account for emails and people (1,000 reveals a day). The same limits apply on the website, the API and MCP.

# no key
curl agentdata.run/api/v1/tech/intercom
curl "agentdata.run/api/v1/companies?sector=Software+%26+SaaS"
curl "agentdata.run/api/v1/signals?days=7"
# free key: contacts
curl -H "Authorization: Bearer $KEY" \
  "agentdata.run/api/v1/lookup?domain=stripe.com"
# or from Claude / Cursor
npx agentdata-mcp-server --api-key $KEY