Pipe catalog
Pipes enrich rows: work emails, mobile numbers, company data, AI steps, and actions like CRM writes or Slack messages. Each one composes with the rest.
Request a pipeFeatured
· Most-used pipes.Find a mobile phone number for any prospect using multiple data providers in a waterfall motion.
Find work email addresses from B2B profile.
Resolve a company to Crustdata's matching candidates by domain or name. Returns the full list of matches.
Find recent company news by performing a search and scrape.
Business description, industry, and financial information.
People Data
21Find work email addresses in a waterfall motion.
Find a standardized JSON representation of a LinkedIn profile.
Find the personal email address of a prospect by using a professional profile URL as input.
Cheaper, less accurate email verification. email:validate:zerobounce@1 is the default email verifier; use this pipe only when the user asks for MillionVerifier or for the cheapest verification.
Find the LinkedIn profile URL by using the prospect's name as input.
Find a mobile phone number for a prospect.
The standard reverse email lookup: email address → the person behind it, as their LinkedIn profile_url plus a full standardized profile (current role and company, work history, education, seniority, location).
Reverse social lookup: a person's Instagram, Facebook, X or TikTok profile URL → who they are, as their LinkedIn profile_url plus a slim identity (name, primary job title, primary company domain, profile picture).
Find one person at a company holding a given job title like "CEO" or "CFO".
The default email verifier. Checks syntax, domain, and mailbox existence, and flags spam traps, abuse, and do-not-mail addresses — the most accurate deliverability verdict in the catalog.
Check if a given email address is personal or professional.
URL only: email address → LinkedIn profile URL, no profile data. For a reverse email lookup use person:identity:email:waterfall@2 (Full LinkedIn profile from email), which returns the same profile_url plus the full profile.
Find every person at a company holding one of the given job titles, as a list. Expand it into rows to enrich each person.
Split a full name into first and last name.
Detect sales signals — job changes, promotions, role changes, work anniversaries, graduations — from a standardized person profile inside a per-row time window.
Combine the first and last names of a prospect into a full name field.
Find list of recent LinkedIn posts.
Check whether the connected Happenstance account's network knows each person — directly or through a friend — and who can introduce you.
Find the recent X (Twitter) posts of an account by handle.
Up to 100 people who reacted to a LinkedIn post, returned inline as a list: name, headline, profile URL and reaction type per person.
Find a developer profile URL (e.g. GitHub) for any prospect using multiple data providers in a waterfall motion.
Company Data
19Find any company's website URL, LinkedIn profile URL, and cleaned name.
Find the other domains a company uses, such as forwarding domains and country sites, each with proof of ownership.
Funding rounds, investors, and total funding in a waterfall motion.
Company data like headcount distribution, LinkedIn URLs etc. via the Crustdata API.
Company data via the Amplemarket API: headcount by department and job function, funding events, technologies, locations and LinkedIn presence. Matches on domain, LinkedIn URL, or both.
Technologies a company's website uses: e-commerce platform, CRM and marketing tools, payments, analytics, ad pixels, hosting and frameworks, grouped by type.
Find companies similar to the company you provide as input.
Extract the company website URL from a work email address.
The creatives a company is running in the Google Ads Transparency Center, with their variations.
Qualify a company on whether it is hiring for a role you care about. Returns its open postings as one list, filtered by function, department, seniority and remote-only. Coverage is partial — many companies return no matching postings — so it suits qualifying rows you already have; to FIND companies hiring, start from a jobs search.
Find who the connected Happenstance account's network knows at each company — current employees, how they are connected, and who can introduce you.
Search LinkedIn posts for every record — posts by a company page, mentioning a company or person, or matching keywords — returned inline as a list.
Fetch the reviews of a local business from its map listing by place_id, sorted by lowest or highest rating, newest or relevance, and optionally narrowed to reviews containing keywords.
Find a local business's map listing from its name and city: place_id, rating and reviews, category, phone, website, address and opening hours.
See which Google searches a website appears for in the normal results: how many, how many on page 1 and in the top 3, estimated visits, and the top keywords with position, search volume and cost per click.
See which Google searches show an AI-written summary that links to a website: how many, how often it is the first link, and the top searches with link position, search volume and cost per click.
Search recent X (Twitter) posts for every record — posts by an account, mentioning it, linking to a domain, or matching keywords — returned inline as a list.
The ads a company is running in the LinkedIn ad library, with their creative.
The ads a company is running in the Meta Ad Library (Facebook, Instagram), with their creative.
Tools
56Extract nested JSON properties into new output fields.
Executes configurable AI prompts with structured output fields.
Answers a yes/no question per row with a decision model and returns a real boolean plus its confidence.
Picks one of your choices per row with a decision model and returns the choice plus its confidence.
Runs an AI research agent with web tools and structured output fields.
Compute a value per row with a JavaScript formula over mapped columns — cleanup, math, conditionals, date arithmetic — without an external provider.
Fetch one web page and return its full content as clean markdown. Handles JavaScript-rendered pages and PDFs.
Transform URL input into a domain. Also tolerates bare domains and email addresses (the part after @ is used). Examples: 'https://go.com' → 'go.com'; 'jane@acme.io' → 'acme.io'.
Expand a list into sheet rows — each item is written as-is to an `expanded_value` column (objects/arrays as json).
Extract relevant content from a website using Parallel's Extract API. Returns a flat list of results — page title, publish date and the excerpts most relevant to your objective.
Make a generic GET/POST/PUT/PATCH/DELETE HTTP request and capture the response — parsed JSON, or the raw text when the body isn't JSON. Optionally validate the response against a JSON schema.
Replaces template tags like {{ name }} with actual values.
An AI agent specialized in composing emails.
Join one or more fields into one field. The final result will be the first defined field. A single-source merge establishes a stable output name that additional sources can join later without consumers changing.
Look up rows in another pipe0 sheet matching a condition block (and/or over eq, neq, gt, gte, lt, lte, contains, not_contains, matches). Outputs the total match count, the matching rows, and their record IDs.
Claim an entry in a bucket (write-if-absent). Wins (is_claim_successful: true) or finds the entry already taken (no_result). Put this before a write pipe to guard against duplicates across sheets, runs, and time.
Write the moment the pipe ran (ISO 8601, UTC) into a column — optionally offset into the future/past or truncated to a date. The building block for freshness cadences, follow-up dates, and expiry columns.
Pause a row for a set time, then let the steps after it run: a follow-up three days after an email, a retry tomorrow.
Check a row on a schedule until a condition holds, at most N times: wait for a job to finish, a reply to arrive, a status to change.
Pause a row until a person approves or rejects it in the sheet. The steps after it read approved: send only what was approved.
Pause a row until a person picks one of your options in the sheet. The steps after it read choice: route each row by the pick.
Count events per entry in an event-counter bucket: each run where the condition matches increments the entry's count; otherwise the current count is read unchanged. Outputs the running count and when it last changed.
Delete rows in ANOTHER pipe0 sheet matching a condition block. The replace-on-refresh building block: delete previously written downstream rows before re-appending fresh ones.
Add rows to a pipe0 sheet in a given project. config.project and config.sheet may each be a literal slug or a `{{ field_name }}` tag.
Rotate through the rows of another sheet (e.g. a sellers roster). Each unique counter value picks the next row round-robin; the selected columns of the picked row become outputs. Re-run safe — the same counter value always picks the same row.
Add this row's identifier to a stream's watch list, so changes about that person arrive as updates. Free; the stream must already exist.
An AI agent specialized in composing messages for modern workplace communication platforms (Slack, Teams, Discord, etc.)
Strip a field value of characters not commonly found in slugs.
Take this row's identifier off a stream's watch list, so it stops being watched and stops costing money. Free.
Assign the next sequence number to an entry in a counter bucket, exactly once (a re-run returns the same number). The building block for round-robin lead routing.
Look up an entry in an existing claim bucket WITHOUT writing. Present → no_result (e.g. suppressed); absent → completed. For suppression / do-not-contact / already-seen gates.
Look up rows in your own PostgreSQL database with a read-only SQL query resolved per row. Outputs the total match count and the matching rows.
Look up rows in your own Databricks SQL warehouse with a read-only SQL query resolved per row. Outputs the total match count and the matching rows.
Find the first row in a Google Sheet that matches a condition on its columns (top or bottom first), and return it keyed by column name, with the number of matching rows. A miss is a normal result, not an error.
Find every row in a Google Sheet that matches a condition on its columns. Outputs the matching rows (keyed by column name, up to a limit) and the total match count.
Find companies in your Attio workspace by domain, alternative domains included. Each exact domain belongs to one company at most; root-domain matching also finds subdomain records.
Find the Email Bison lead for an email address, with its status in each campaign. A clean miss completes with found = false.
Find the Nooks prospect for an email address. Nooks prospects are synced from your CRM; a clean miss completes with found = false.
Find the Nooks user (a seat in your workspace) for an email address — e.g. to route each enrollment to the right owner.
Check whether a contact exists in your HubSpot portal by email. Matches the primary email exactly (and optionally secondary emails); a clean miss completes with found = false.
Find companies in your HubSpot portal by domain, alternative domains included. Domains aren't unique in HubSpot, so 0..N matches are normal. Each match carries its contact and deal counts.
Find deals in your HubSpot portal for a company (by domain) or contact (by email) through HubSpot's associations. Deals have no natural key of their own.
Find records of any HubSpot object by any property: custom domain fields, LinkedIn URLs, external IDs. Returns the count and up to 25 matches, oldest or newest first.
Retrieve the HubSpot records associated with a source record — one level of joins across standard and custom objects (e.g. a company's deals, or a custom object linked to a company).
Check whether a person exists in your Salesforce org by email — across Contacts, Leads, and Person Accounts in one call, with deterministic precedence and converted-lead redirect. A clean miss completes with found = false.
Find accounts in your Salesforce org by company domain, alternative domains included. The match probes exact spellings of the Website URL, never substrings. Multiple matches are normal. Optionally counts each match's contacts and opportunities.
Find opportunities in your Salesforce org for a company (by domain, through its account) or a person (by email, through contact roles). Opportunities have no natural key of their own. A clean miss completes with count = 0.
Fetch a full Salesforce record (any standard or custom object) by its record ID — every field visible to the connected user. A deleted or missing record completes with found = false.
Find records of any Salesforce object by any field: custom domain fields, domain-map objects, LinkedIn URLs, external IDs, DuplicateRecordItem. Returns the count and up to 25 matches, oldest or newest first.
Read a Salesforce record's activity timeline — tasks and events (calls, emails, meetings) associated with a contact, lead, account, or opportunity. Returns the true activity count, the last-activity date, and the timeline items. An empty timeline completes with count = 0.
Check whether a person exists in your Attio workspace by email. email_addresses is Attio's unique identity attribute, so any address on a record matches exactly; a clean miss completes with found = false.
Find deals in your Attio workspace linked to a company (by domain) or person (by email), optionally narrowed to specific stages. Deals have no natural key — matching traverses the relationship. A clean miss completes with a count of 0.
Find one record of ANY Attio object (custom objects included) by a unique attribute — the generic find-by-key. A clean miss completes with found = false.
Fetch one Attio record of any object (custom objects included) by its record ID — the full record with every attribute. A missing record completes with found = false.
Find records of any Attio object by any attribute: custom domain fields, LinkedIn URLs, external IDs. Returns the count and up to 25 matches, oldest or newest first.
Read the activity timeline of one Attio record — notes, meetings, and tasks linked to it — merged newest-first. A record with no activity completes with found count = 0.
Actions
44Send a Slack message to a channel.
Create a draft email in Gmail (does not send).
Add a lead to an Amplemarket sequence, with custom dynamic fields.
Add a lead to an Instantly campaign, with custom variables.
Send an email with Gmail.
Create a Google Calendar event per record, with attendees, an optional Google Meet link, and invitations sent by Google.
Send an email with Resend.
Create a contact inside a Resend audience.
Add a lead to a lemlist campaign sequence, with custom variables.
Enroll a Nooks prospect into a Nooks sequence.
Create an Email Bison lead, or update the existing one with the same email, with its custom variables.
Add an Email Bison lead to an Email Bison campaign.
Send a Slack direct message to a workspace member.
Publish a post from your connected X (Twitter) account.
Update a HubSpot record (standard or custom object) by its record ID, mapping columns to properties. Blank values are skipped by default so un-enriched cells never erase CRM data.
Delete (archive) a HubSpot record by its record ID: a contact, company, deal, ticket, engagement or custom object. Deleted records stay restorable in HubSpot for up to 90 days.
Merge a duplicate HubSpot record into the record you keep: a contact, company, deal, ticket or custom object. HubSpot can't undo a merge.
Log an activity in HubSpot — a note, task, call, email, or meeting engagement — and associate it with a contact, company, deal, or custom object so it shows on the record's timeline.
Create-or-update a HubSpot contact by email. One match updates, none creates, several fail and write nothing; `created` says which happened.
Create-or-update any HubSpot record (standard or custom object) by a property HubSpot marks unique. One match updates, none creates; `created` says which happened.
Create-or-update a HubSpot company by domain, alternative domains included. One match updates, none creates, several fail and write nothing; `created` says which happened.
Make sure two HubSpot records are linked — e.g. a contact to its company — with a chosen association label. Never removes a link or label; re-runs are no-ops.
Remove the link between two HubSpot records — e.g. a contact and a company it no longer belongs to — with every label on it. Touches no other link; removing a link that doesn't exist succeeds.
Create a HubSpot deal in a chosen pipeline and stage, linked to a company or contact in the same call. Always creates — every run makes a new deal.
Update a Salesforce record (standard or custom object) by its record ID, mapping columns to fields. Blank values are skipped so un-enriched cells never erase CRM data; the output carries the fresh post-save record.
Delete a Salesforce record by its record ID: a lead, contact, account, opportunity or custom object. Salesforce keeps it in the Recycle Bin for 15 days.
Merge a duplicate Salesforce account, contact or lead into the record you keep. Related records move over; the duplicate stays in the Recycle Bin for 15 days.
Find a Salesforce record by an External ID field and create it when it's missing; existing records are left unchanged or updated. `created` says which happened.
Find a Salesforce contact by email and create it when it's missing; existing contacts are left unchanged or updated. `created` says which happened.
Find an open Salesforce lead by email and create it when it's missing; existing leads are left unchanged or updated. `created` says which happened.
Find a Salesforce account by company domain, alternative domains included, and create it when it's missing; existing accounts are left unchanged or updated. `created` says which happened.
Create a Salesforce opportunity in a chosen stage, linked to its account through AccountId.
Log an activity in Salesforce as a Task (the native 'Log a Call' idiom — subtype call/email/task) against a contact, lead, account, or opportunity. Optionally rerun-safe via an External ID idempotency key.
Update an Attio record (standard or custom object) by its record ID, mapping fields to attributes. Blank values are skipped so un-enriched cells never erase CRM data.
Permanently delete an Attio record by its record ID: a person, company, deal or custom object. Attio has no restore.
Merge a duplicate Attio record into the record you keep: a person, company, deal or custom object. Attio can't undo a merge.
Append each record as a new row to a Google Sheet, mapping fields to the tab's columns. Rows are written in batches, as plain text.
Update every Google Sheets row that matches a condition on its columns, writing mapped fields into the tab's columns. Optionally add a row when nothing matches.
Find an Attio person by email and create them when they're missing; existing people are left unchanged or updated. `created` says which happened.
Find an Attio company by domain, alternative domains included, and create it when it's missing; existing companies are left unchanged or updated. `created` says which happened.
Find an Attio record of any object by a unique attribute and create it when it's missing; existing records are left unchanged or updated. `created` says which happened.
Create an Attio deal in a chosen stage with an owner, linked to its company and people.
Add a record to an Attio list (pipeline) unless it's already there; its existing entry is left unchanged or updated, e.g. to move its stage. `created` says which happened.
Take a record off an Attio list (pipeline): deletes its entry with the entry's list attributes. The record itself stays.
