Email Finder: Find Work Emails by Name & Company Domain

Give it a first name, a last name and a company domain and get the most likely work email addresses, ranked by evidence, plus domain checks: syntax, MX records, disposable domains, free providers and typo suggestions. Export JSON/CSV/Excel or call it from an AI agent.

Read this first: there is NO SMTP check. This Actor never connects to a mail server to ask if a mailbox exists. It does not verify that a mailbox exists, and it does not detect catch-all domains. It tells you which address patterns are plausible and whether the domain can receive mail. Outbound SMTP (port 25) is blocked on the Apify platform, so a mailbox-level check is not possible here. If you need mailbox verification, run the candidates through a dedicated verifier before sending.

🇪🇸 Español: encuentra el correo laboral más probable a partir de nombre, apellido y dominio de la empresa. Genera patrones (nombre.apellido, inicial+apellido, etc.), los ordena según evidencia publicada en el sitio de la empresa y valida el dominio (sintaxis, MX, dominios desechables, errores tipográficos). No usa SMTP: no confirma que el buzón exista ni detecta dominios catch-all. 🇧🇷 Português: encontra o e-mail corporativo mais provável a partir de nome, sobrenome e domínio da empresa. Gera padrões (nome.sobrenome, inicial+sobrenome etc.), ordena por evidências publicadas no site da empresa e valida o domínio (sintaxe, MX, domínios descartáveis, erros de digitação). Não usa SMTP: não confirma se a caixa postal existe nem detecta domínios catch-all.

What you get

• 🔎 Ranked email candidates per person (up to 11 patterns: first.last, flast, first, firstlast, f.last, first_last, firstl, last, last.first, lastf, first-last).
• 🏷️ A confidence label per candidate, based on what the company's own website shows (see below). It is a ranking label, not a probability: we publish no accuracy percentage.
• ✅ Domain verification: valid syntax, MX records (and the mail host), disposable-domain blocklist (about 9,900 domains), free providers (gmail.com...), typo suggestion (gmial.com → gmail.com, .con → .com).
• 📨 Email checker mode: pass addresses in emailsToCheck to get syntax, MX, disposable, free-provider, role-based (info@, sales@) and typo flags.
• 🌎 Accents and compound surnames are handled (María González → maria.gonzalez); works with any domain, including .com.br, .mx, .co.
• 💸 $3 per 1,000 contacts, plus $0.005 per run. Runs with no results cost nothing.

How the confidence label works

confidence	evidence	Meaning
high	seen_on_website	This exact address is published on the company website (home, contact, about, team pages).
medium	matches_site_convention	At least 2 person addresses published on the site use the same pattern (for example two first.last addresses).
low	single_site_example	One address on the site has this pattern.
low	common_pattern	Nothing on the site; this is only a common pattern, in a fixed conventional order.

Most results will be low: many companies do not publish staff addresses. In our tests (Oct 24, 2026) we got high for a person listed on a public NASA page and low for most commercial sites (HubSpot, Stripe, Microsoft, Shopify and Mercado Libre showed no person addresses on the pages we read). Treat low as a guess to verify elsewhere. The siteConvention field shows what the site published, so you can judge the evidence yourself.

Sources and coverage

Source	Coverage	Data
Company website	Any domain	emails published on the site, used to learn the company's email pattern
DNS (MX records)	Any domain	whether the domain accepts email, mail provider
Pattern engine	Any name + domain	ranked candidates (first.last@, flast@...) with confidence label

How to use it

1. Click Try for free (or Start): the prefilled input runs a quick test.
2. Fill in the input fields (see the examples below). Every field is optional.
3. Click Start and wait a few seconds to minutes, depending on how many contacts you ask for.
4. Open the Output tab and export as JSON, CSV, Excel, XML or HTML, or read the results via API.

Input examples

Two contacts (default):

{"contacts": [{"firstName": "Jane", "lastName": "Doe", "domain": "hubspot.com"}]}

Bigger list, top 5 candidates:

{"contacts": [{"firstName": "John", "lastName": "Smith", "domain": "stripe.com"}, {"firstName": "Ana", "lastName": "García", "domain": "mercadolibre.com"}], "candidatesPerContact": 5}

Only check email domains:

{"emailsToCheck": ["sales@acme.com", "jane@gmial.com"]}

Fast mode (no website scan):

{"contacts": [{"firstName": "Jane", "lastName": "Doe", "domain": "notion.so"}], "scanWebsite": false}

Input fields

Field	Type	Description
contacts	array	List of people to find. Each item is {"firstName","lastName","domain"} (or {"fullName","domain"}), or a string like "Jane Doe, hubspot.com". Leave empty and fill only …
emailsToCheck	array	Optional addresses to validate at domain level: syntax, MX, disposable, free provider, role-based, typo suggestion. It does NOT check that the mailbox exists.
candidatesPerContact	integer	How many ranked email candidates to return per person (the best goes in 'email', the rest in 'alternatives').
scanWebsite	boolean	Reads a few pages of the company site (home, contact, about, team) to find published addresses. They raise the confidence of a matching pattern. Turn off for faster, p…
maxPagesPerDomain	integer	Pages to read per company domain when scanning the website.
includeFailedRows	boolean	Also output contacts whose domain is invalid, has no MX or is disposable, with the reason. Those rows are never charged.
maxResults	integer	Hard cap on charged rows for the whole run. Never exceeded.

Output example

One row per item. Values below are example values from a real test run (long text shortened):

{
  "mode": "find",
  "firstName": "jane",
  "lastName": "doe",
  "domain": "hubspot.com",
  "input": {"firstName": "Jane", "lastName": "Doe", "domain": "hubspot.com"},
  "email": "jane.doe@hubspot.com",
  "emailPattern": "first.last",
  "confidence": "low",
  "evidence": "common_pattern",
  "alternatives": ["{…}", "{…}"],
  "smtpVerified": false,
  "verificationNote": "Domain-level checks only (syntax, MX, disposable). Mailbox existence and catch-all were NOT checked.",
  "domainStatus": "deliverable_domain",
  "domainHasMx": true,
  "mxHost": "smtp.google.com",
  "isDisposableDomain": false,
  "isFreeProvider": false,
  "websitePagesScanned": 5
  "…": "more fields below"
}

Fields a source does not publish are null.

Output fields

Field	Description
mode	find (name + domain) or check (email validation)
firstName	First name used (ASCII, lowercase)
lastName	Last name used (ASCII, lowercase, spaces removed)
domain	Company domain (normalized, no www)
input	Your original input item
email	Best candidate email (rank 1), or the address checked in 'check' mode
emailPattern	Pattern of the best candidate (first.last, flast, first...)
confidence	high = this exact address is published on the company website; medium = at least 2 person addresses published on the site use this pattern; low = single example or only a common pattern, no real evidence. A ranking label, NOT a probability
evidence	seen_on_website (this exact address is published on the site), matches_site_convention (2+ person addresses on the site share this pattern), single_site_example (1 address on the site has this pattern) or common_pattern (no evidence)
alternatives	Other ranked candidates: email, pattern, evidence, confidence
smtpVerified	Always false: this Actor never uses SMTP and does not verify mailboxes or catch-all
verificationNote	Plain-language limit of the verification
domainStatus	deliverable_domain (has MX), no_mx_a_record_only, domain_not_found, invalid_domain_syntax or invalid_syntax
domainHasMx	True when the domain publishes MX records
mxHost	Primary mail server host
isDisposableDomain	True when the domain is in a public disposable-email blocklist
isFreeProvider	True for free providers such as gmail.com
domainTypoSuggestion	Likely correct domain when the input looks like a typo (gmial.com, .con)
websitePagesScanned	Pages of the company website read
websiteError	Why the site could not be read, if so
siteConvention	Person-address patterns published on the site: counts, dominant pattern and sample local parts
scrapedAt	ISO timestamp

The Output tab has a Contacts view.

Pricing

Pay-per-event: you pay only for results delivered, no monthly fee.

• Contact with email candidates (contact): $3.00 per 1,000 ($0.003 each)
• Email domain check (email-check): $1.00 per 1,000 ($0.001 each)
• Actor start: $0.005 per run

Run	Contacts	Cost
Quick test	20	$0.065
Small run	100	$0.305
Regular run	1,000	$3.005
Large run	10,000	$30.005

Rows that are not delivered are never charged. Apify's free plan includes monthly credit, and you can set a maximum cost per run in the run options.

Use cases

Outbound sales prospecting · recruiting outreach · PR and link-building lists · cleaning lead lists (typos, disposable and invalid domains) · feeding AI agents that need a work-email candidate for a contact.

Integrations and API

• Export results as JSON, CSV, Excel, XML or HTML from the Output tab.
• Schedule runs (daily, hourly) and send results to Google Sheets, Slack, Zapier, Make, n8n, Airbyte or webhooks.
• Call it from code with the Apify API:

curl -X POST "https://api.apify.com/v2/acts/datadesk-io~email-finder-by-name-domain/run-sync-get-dataset-items?token=YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"contacts": [{"firstName": "Jane", "lastName": "Doe", "domain": "hubspot.com"}]}'

from apify_client import ApifyClient
client = ApifyClient("YOUR_TOKEN")
run = client.actor("datadesk-io/email-finder-by-name-domain").call(run_input={'contacts': [{'firstName': 'Jane', 'lastName': 'Doe', 'domain': 'hubspot.com'}]})
for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(item.get('domain'), item.get('email'))

FAQ

Why is nearly everything low? Because most company sites do not publish staff emails. The label tells you honestly when there is no evidence. Can I send to the candidates directly? Verify them first with a mailbox verifier; wrong guesses can bounce and hurt your sender reputation. Issues? Report them on the Issues tab. We usually reply within 48 h.

Do I need a login, API key or proxies? No. It uses publicly available data only, so no account on the source site is needed.

How fresh is the data? Live: every run reads the source at that moment. Schedule it to keep a dataset up to date.

Why did I get fewer contacts than the maximum? The source had fewer matching contacts, or duplicates were removed. You only pay for contacts delivered.

Limitations

• No SMTP, no mailbox verification, no catch-all detection. deliverable_domain means the domain publishes MX records, not that the person's mailbox exists.
• Candidates are guesses based on patterns. A high label means the address is published on the company's site, which is strong evidence but not a delivery guarantee.
• The website scan reads a few public pages with a plain HTTP client. Sites that block bots or render everything with JavaScript may give no evidence (websiteError tells you).
• The tool does not search LinkedIn, social networks or leaked databases, and it never uses personal data you did not provide.
• Email addresses are personal data in many jurisdictions. Use the results lawfully (GDPR, CAN-SPAM, LGPD) and respect opt-outs.

Support

Found a bug, or need a new field, source or filter? Open an issue in the Issues tab; we usually answer within 24 hours. Custom data work: datadesk.io@proton.me.

Use with AI agents (MCP)

This Actor works as a tool for AI agents through the Apify MCP server. Add it to Claude, ChatGPT, Cursor, VS Code or any MCP client:

https://mcp.apify.com/?tools=datadesk-io/email-finder-by-name-domain

Sign in with OAuth or Authorization: Bearer <APIFY_TOKEN>. Local stdio config:

{"mcpServers": {"email-finder": {"command": "npx", "args": ["-y", "@apify/actors-mcp-server", "--tools", "datadesk-io/email-finder-by-name-domain"], "env": {"APIFY_TOKEN": "<token>"}}}}

What goes in → what comes out: Input: a list of people (first name, last name, company domain) and/or email addresses to check. Output: one JSON row per person with the best email candidate, ranked alternatives, a confidence label with its evidence, and domain checks (MX, disposable, typo). Every input is optional and every output field is documented. The agent should treat results as candidates, not verified mailboxes (smtpVerified is always false).

Example prompt: "Find the likely work email for Maria Garcia at shopify.com and tell me how confident the match is." Use maxResults to cap the cost of one call.