Kaggle Email Scraper

Kaggle Email Scraper - extract public data scientist and ML practitioner emails from Kaggle

The Kaggle Email Scraper collects publicly indexed contact emails from Kaggle: data scientists, machine learning engineers, notebook authors, dataset publishers and competition participants who left an address in public text.

You supply keywords, an optional location and a list of email domains. The Kaggle Email Scraper builds domain-scoped Google queries against kaggle.com, parses each result block and writes every matching address into a structured dataset.

There is no login, no Kaggle API, no headless browser and no cookies. This Kaggle email extractor reads only what Google has already indexed publicly.

One honest caveat up front. Kaggle yields less cleanly than creator platforms, because a large share of indexed kaggle.com pages are datasets, competitions and notebooks rather than user profiles, so a fair proportion of rows arrive with an email but no username and no profile URL.

That is worth knowing before you plan a run. The addresses are still real and usable; you simply get fewer rows where a handle and a kaggle.com profile URL come attached.

Technical recruiters, ML tooling vendors, developer-relations teams, bootcamps and research groups use the Kaggle Email Scraper for contact discovery across the data science community.

---

Features of the Kaggle Email Scraper

Everything below is implemented in the Kaggle Email Scraper today. No roadmap items, no aspirational claims.

Feature	What it means in practice
Google site: search automation	Every query is scoped to kaggle.com, so results stay on-platform
Query expansion	Base, quoted and intitle: variants plus one variant per query modifier
Recruiting-tuned modifiers	Defaults are email, contact, inquiries, hire, work with me
Domain-filtered extraction	Only emails on your customDomains list are kept
Global deduplication	Each unique address reaches the dataset exactly once across all queries and pages
Obfuscation-aware parser	Understands name [at] domain [dot] com, name (at) domain, name @ domain.com, domain .com
Unicode normalisation	Handles zero-width characters and the full-width at sign
Junk filter	Rejects placeholders such as email@, yourname@, test@, xxx@ and single-character locals
Boundary-correct matching	@gmail.com will not match inside @gmail.company or @gmail.com.br
Soft-wrap repair	Discards a hit that is only the tail of another email in the same result block
Concurrency control	An asyncio worker pool runs queries in parallel with a shared stop signal on maxEmails
Retry logic	Up to 3 attempts per page with exponential backoff and a fresh proxy session per request
Block detection	CAPTCHA, "unusual traffic" and consent pages are detected and retried, never scored as empty
Re-queue pass	Blocked or failed queries are retried once more at the end of the run
Resumable state	Progress persists in the key-value store, keyed by a hash of your input
Structural parsing	Result blocks are located via the <h3> title, not by Google's CSS class names
Fallback parser	A Google markup change degrades to "emails without account details", not "no emails"
Streaming dataset writes	Leads appear in the dataset immediately, so exports can begin mid-run

---

How the Kaggle Email Scraper works

The Kaggle Email Scraper is a search-results crawler. It never opens kaggle.com, never downloads a dataset and never executes a notebook.

1. Read input. Keywords, location, email domains, page caps and concurrency are validated first.

2. Build queries. Queries use the site: operator, for example: site:kaggle.com machine learning engineer "@gmail.com" "Bangalore".

3. Expand queries. With expandQueries on, each keyword and domain pair is issued as a base query, a quoted variant, an intitle: variant and one variant per entry in queryModifiers. Base queries always run first.

4. Fetch pages. The Kaggle Email Scraper fetches result pages asynchronously with aiohttp through the Apify GOOGLE_SERP proxy, paginating up to maxPagesPerQuery.

5. Parse result blocks. The parser locates the <h3> title, then the smallest surrounding block, so titles, snippets and site labels stay attached to the correct record.

6. Extract and normalise. A domain-filtered regex pulls candidates from the block text, normalises them, and removes junk, tails and boundary mismatches.

7. Deduplicate and store. Every new address is deduplicated globally and pushed straight into the Apify dataset as structured data.

The Kaggle Email Scraper run summary logs pages fetched, blocked pages, retries and emails per page, which is especially useful here for comparing profile-shaped keywords against notebook-shaped ones.

---

What data does the Kaggle Email Scraper extract?

Each row is one email paired with whatever metadata Google exposed alongside it.

You get the address, the matched domain, the account or page label, a parsed name where the title looks like a profile, a handle and profile URL where Kaggle exposes one, a cleaned snippet and full provenance.

Provenance matters. The keyword and query fields tell you which phrasing surfaced this record, which is how you steer the next Kaggle Email Scraper run toward profile pages rather than dataset pages.

When a handle is available the Kaggle Email Scraper rebuilds profileUrl as https://www.kaggle.com/<handle>. When Google returns a dataset, competition or notebook page instead, username will be null and profileUrl empty.

---

Input schema for the Kaggle Email Scraper

Every Kaggle Email Scraper input field is listed below with the exact name, type and default from the Actor input schema.

Field	Type	Default	Description
keywords	array (required)	["data scientist", "machine learning"]	Search terms describing the Kaggle accounts you want (niche, job title, industry)
location	string	""	Optional location phrase added to every query, e.g. "New York"
customDomains	array	["@gmail.com", "@yahoo.com"]	Only emails ending with one of these domains are collected; leading @ optional
maxEmails	integer 1-10000	20	Stop once this many unique emails have been collected
countryCode	string	""	Two-letter country code for the search proxy (US, GB, DE). Empty for any
expandQueries	boolean	true	Search each keyword and domain pair with several phrasings
queryModifiers	array	["email", "contact", "inquiries", "hire", "work with me"]	Extra words combined with each keyword when expansion is on
maxPagesPerQuery	integer 1-50	30	Page cap per query
maxConcurrency	integer 1-20	5	How many queries run in parallel

Input example

{
  "keywords": ["data scientist", "machine learning engineer", "computer vision"],
  "location": "",
  "customDomains": ["@gmail.com", "@outlook.com"],
  "maxEmails": 300,
  "countryCode": "US",
  "expandQueries": true,
  "queryModifiers": ["email", "contact", "inquiries", "hire", "work with me"],
  "maxPagesPerQuery": 30,
  "maxConcurrency": 5
}

---

Output schema of the Kaggle Email Scraper

Every dataset item the Kaggle Email Scraper produces carries all fourteen fields below.

Field	Meaning
network	Platform name
keyword	The keyword that produced the lead
query	The exact Google query used
title	Raw result title
accountName	Account label Google prints (handle, display name or page label)
fullName	Display name parsed from a profile-style title; empty for dataset and notebook titles
username	URL-safe handle when Kaggle exposes one; otherwise null
profileUrl	Canonical account URL when a handle is known; otherwise empty
url	Direct Kaggle link when exposed, else the profile URL
description	Bio or page snippet, cleaned of labels and engagement counters
email	Lower-cased email address
emailDomain	The matched domain, e.g. @gmail.com
possiblyTruncated	true when Google's snippet ellipsis touched the email - verify before sending
foundAt	ISO 8601 UTC timestamp

Output example

{
  "network": "Kaggle",
  "keyword": "computer vision",
  "query": "site:kaggle.com computer vision \"@gmail.com\" contact",
  "title": "Ravi Deshmukh | Notebooks Expert | Kaggle",
  "accountName": "ravideshmukh",
  "fullName": "Ravi Deshmukh",
  "username": "ravideshmukh",
  "profileUrl": "https://www.kaggle.com/ravideshmukh",
  "url": "https://www.kaggle.com/ravideshmukh",
  "description": "Computer vision and medical imaging. Open to consulting and research collaborations. Contact: ravi.deshmukh.ml@gmail.com",
  "email": "ravi.deshmukh.ml@gmail.com",
  "emailDomain": "@gmail.com",
  "possiblyTruncated": false,
  "foundAt": "2026-08-31T14:12:09.348Z"
}

The Kaggle Email Scraper dataset exports as JSON, CSV, Excel or XML from the Apify console, or streams through the Apify API into your ATS, CRM or outreach tool.

---

How to use the Kaggle Email Scraper

Step 1 - bias your keywords toward people, not projects. Terms like data scientist, machine learning engineer, NLP researcher or Kaggle Master bring back more profile pages; terms like titanic dataset bring back competition pages with no owner attached.

Step 2 - pick sensible email domains. @gmail.com dominates among individual practitioners. Add academic or corporate domains if you are targeting a specific institution.

Step 3 - start small. Run the Kaggle Email Scraper with maxEmails at 20 to 50 first, then inspect how many rows came back with a populated username.

Step 4 - keep the recruiting-tuned modifiers. hire, contact, inquiries and work with me match the phrasing people use when they are open to work.

Step 5 - add location for geography-bound hiring. Practitioners sometimes state a city or country, but many do not, so expect location filtering to reduce volume noticeably.

Step 6 - clean before you send. Filter out rows where possiblyTruncated is true, split rows with and without a username, then run the remaining Kaggle Email Scraper results through email verification.

---

Use cases for the Kaggle Email Scraper

Use case	Who runs it	Typical keywords
Data science and ML recruiting	Technical recruiters and in-house sourcers	data scientist, machine learning engineer, kaggle master
Developer tools and MLOps prospecting	Founders and sales teams at ML vendors	deep learning, pytorch, feature engineering
Developer relations and community building	DevRel and community managers	notebook, tutorial, open source
Research collaboration outreach	Labs, universities and research groups	nlp researcher, computer vision, time series
Bootcamp and course enrolment	Education providers and training platforms	beginner, learning data science, portfolio
Freelance and contract data work	Agencies and consultancies	freelance data scientist, analytics consultant
Conference and hackathon recruiting	Event organisers and sponsors	competition, hackathon, benchmark
Talent-market research	Analysts studying the ML labour market	any role keyword, varied by location

Kaggle is a demonstrated-skill community rather than a resume site, which is why sourcers value it, and why the Kaggle Email Scraper is best used as one input among several rather than a standalone list.

For academic contacts the ResearchGate Email Scraper and the Academia.edu Email Scraper are usually richer, and practitioners who publish courses often appear in the Coursera Email Scraper or the Pluralsight Email Scraper.

---

Why choose this Kaggle Email Scraper

Manual sourcing on Kaggle means paging through leaderboards, opening profiles, reading bios and copying addresses. The Kaggle Email Scraper automates that loop and returns a deduplicated dataset.

Every row is auditable. The query and keyword fields show exactly how a lead was found, which matters more here than elsewhere because it lets you separate profile-shaped queries from dataset-shaped ones.

The extraction layer is deliberately conservative. Junk filtering, boundary-correct domain matching and soft-wrap repair mean far fewer garbage rows to clean downstream.

Long Kaggle Email Scraper runs survive interruption. State persists in the key-value store and saves fire on Apify's PERSIST_STATE, MIGRATING and ABORTING events.

---

Worked example: a computer vision hiring pipeline

You are hiring three computer vision engineers and want a top-of-funnel list of 300 practitioners.

Set keywords to ["computer vision", "image segmentation", "kaggle master"], customDomains to ["@gmail.com"], countryCode to "US" and maxEmails to 300, leaving expandQueries on.

The Kaggle Email Scraper issues base, quoted and intitle: variants plus one query per modifier, paginating each independently and streaming leads as they are found.

Export the Kaggle Email Scraper dataset to CSV, drop possiblyTruncated rows, then sort by whether username is populated so your sourcers work the identifiable profiles first.

---

Limitations you should know before running

The Kaggle Email Scraper is deliberately transparent about its boundaries. Read these before planning a large data collection run.

Weaker identity resolution than creator platforms. Many indexed kaggle.com pages are datasets, competitions and notebooks rather than user profiles, so a fair share of Kaggle Email Scraper rows carry an email with no handle attached.

Only publicly indexed emails. The Kaggle Email Scraper returns addresses that already appear in Google's index. Practitioners who never published an address will not surface.

Google's per-query ceiling. A single query caps out at roughly 300 results. Query expansion exists to work around that ceiling, which is why it defaults to on.

possiblyTruncated. When Google's snippet ellipsis touches an address, this field is true and the email may be incomplete. Verify those rows before sending.

Apify GOOGLE_SERP proxy required. The Kaggle Email Scraper cannot run without Apify proxy credentials that include the GOOGLE_SERP group.

Free-plan cap. Running the Kaggle Email Scraper on a free Apify plan limits results to 100 emails per run. Paid plans are uncapped.

Handle availability. username and profileUrl are populated only when Google's result exposes a Kaggle handle. Rows without one still carry accountName and, sometimes, fullName. This is a Google limitation, not a bug.

No guaranteed volume. Kaggle Email Scraper yield varies with keywords, domains and location. No result count is promised.

The Kaggle Email Scraper is not affiliated with, endorsed by or officially supported by Kaggle.

---

Related Actors

Every Actor below runs the same engine as the Kaggle Email Scraper, with identical input fields and an identical output schema; only the target platform differs.

Actor	What it collects
Kaggle Email and Phone Number Scraper	Emails and phone numbers from Kaggle
Kaggle Phone Number Scraper	Public phone numbers from Kaggle
Academia.edu Email Scraper	Public contact emails from Academia.edu
Buy Me a Coffee Email Scraper	Public contact emails from Buy Me a Coffee
Carrd Email Scraper	Public contact emails from Carrd
Domestika Email Scraper	Public contact emails from Domestika
edX Email Scraper	Public contact emails from edX
Gumroad Email Scraper	Public contact emails from Gumroad
Kajabi Email Scraper	Public contact emails from Kajabi
Ko-fi Email Scraper	Public contact emails from Ko-fi
Linktree Email Scraper	Public contact emails from Linktree
MasterClass Email Scraper	Public contact emails from MasterClass
Meetup Email Scraper	Public contact emails from Meetup
Pluralsight Email Scraper	Public contact emails from Pluralsight
Preply Email Scraper	Public contact emails from Preply
ResearchGate Email Scraper	Public contact emails from ResearchGate
Skillshare Email Scraper	Public contact emails from Skillshare
Teachable Email Scraper	Public contact emails from Teachable
Udemy Email Scraper	Public contact emails from Udemy
Academia.edu Email and Phone Number Scraper	Emails and phone numbers from Academia.edu
Buy Me a Coffee Email and Phone Number Scraper	Emails and phone numbers from Buy Me a Coffee
Carrd Email and Phone Number Scraper	Emails and phone numbers from Carrd
Domestika Email and Phone Number Scraper	Emails and phone numbers from Domestika
edX Email and Phone Number Scraper	Emails and phone numbers from edX
Gumroad Email and Phone Number Scraper	Emails and phone numbers from Gumroad
Kajabi Email and Phone Number Scraper	Emails and phone numbers from Kajabi
Ko-fi Email and Phone Number Scraper	Emails and phone numbers from Ko-fi
Linktree Email and Phone Number Scraper	Emails and phone numbers from Linktree
MasterClass Email and Phone Number Scraper	Emails and phone numbers from MasterClass
Meetup Email and Phone Number Scraper	Emails and phone numbers from Meetup
Pluralsight Email and Phone Number Scraper	Emails and phone numbers from Pluralsight
Preply Email and Phone Number Scraper	Emails and phone numbers from Preply
ResearchGate Email and Phone Number Scraper	Emails and phone numbers from ResearchGate
Skillshare Email and Phone Number Scraper	Emails and phone numbers from Skillshare
Teachable Email and Phone Number Scraper	Emails and phone numbers from Teachable
Academia.edu Phone Number Scraper	Public phone numbers from Academia.edu
Buy Me a Coffee Phone Number Scraper	Public phone numbers from Buy Me a Coffee
Carrd Phone Number Scraper	Public phone numbers from Carrd
Domestika Phone Number Scraper	Public phone numbers from Domestika
edX Phone Number Scraper	Public phone numbers from edX
Gumroad Phone Number Scraper	Public phone numbers from Gumroad
Kajabi Phone Number Scraper	Public phone numbers from Kajabi

FAQ about the Kaggle Email Scraper

What is the Kaggle Email Scraper?

It is an Apify Actor that extracts publicly indexed contact emails from Kaggle by running domain-scoped Google searches and parsing the resulting search-result blocks.

Does the Kaggle Email Scraper log into Kaggle or use its API?

No. It never logs in, never calls a Kaggle API and never opens a browser, so there are no cookies and no authentication anywhere in the pipeline.

Why do so many rows have no username?

Because much of Kaggle's indexed surface is datasets, competitions and notebooks rather than profiles. Those pages have no handle for the Kaggle Email Scraper to resolve, so username is null and profileUrl is empty.

Can I improve the share of rows with a handle?

Often, yes. Role-shaped keywords such as data scientist, machine learning engineer or kaggle master return more profile pages than dataset-shaped keywords.

Where do the emails actually come from?

From publicly indexed titles, snippets and site labels, typically bios and descriptions where someone published an address themselves.

What does possiblyTruncated: true mean?

Google's snippet ellipsis touched the address, so it may be cut off. Verify those rows before contacting anyone.

How many emails will one run of the Kaggle Email Scraper return?

maxEmails accepts 1 to 10000, but free Apify plans are capped at 100 emails per run and paid plans are uncapped. Real volume depends on keywords, domains and location.

Is a proxy required?

Yes. The Kaggle Email Scraper requires the Apify GOOGLE_SERP proxy and cannot run without Apify proxy credentials.

Should I leave expandQueries on?

Almost always. Google returns roughly 300 results per query, and expansion multiplies your reachable surface through base, quoted, intitle: and modifier variants.

Can I target a specific country?

Yes. Set countryCode for the search proxy and optionally add a location phrase, keeping in mind that many practitioners never state a city.

What happens if my run is interrupted or migrated?

State persists in the key-value store keyed by a hash of your input, with throttled saves plus saves on PERSIST_STATE, MIGRATING and ABORTING, so an interrupted run resumes instead of restarting.

Can I email everyone in the dataset immediately?

Treat the Kaggle Email Scraper output as raw lead data. Deduplication is automatic, but verification, consent and compliance with GDPR, CAN-SPAM and Kaggle's terms remain your responsibility.

---

Leave a review

If the Kaggle Email Scraper saved you time, please leave a star rating and a short review on the Actor page.

Reviews are how other buyers judge whether a tool works, and they tell us which features to build next.

If something did not work, email neurodata.apify@gmail.com instead - bugs get fixed faster than they get complained about.

Support

Questions about the Kaggle Email Scraper, bug reports or a custom build? Email neurodata.apify@gmail.com.