Greenhouse

This documentation for the Greenhouse integration describes the technical capabilities of this integration, including authorization, scopes/permissions, and utilized endpoints. For more information on how to integrate Greenhouse, visit our connection instructions.

Version

This integration utilizes the Greenhouse Harvest API v3.

Base URL

The base URL used for all Greenhouse API endpoints:

https://harvest.greenhouse.io/v3

---

Authentication & Authorization

The DataGrail Greenhouse integration connects using OAuth 2.0 with the following credentials: Client ID and Client Secret.

Sensitive Credentials

Publicly exposing your API credentials can allow unauthorized access to Greenhouse API endpoints by a third party. DataGrail stores your API credentials encrypted and protected.

Scopes

The Greenhouse integration requires specific scopes that must be granted in order to function for a given capability.

Scope	Base	Access	Deletion
harvest:candidates:list	✅	✅	✅
harvest:applications:list		✅
harvest:jobs:list		✳️
harvest:rejection_reasons:list		✳️
harvest:offers:list		✳️
harvest:attachments:list		✳️
harvest:users:list		✳️
harvest:sources:list		✳️
harvest:candidates:destroy			✅
harvest:candidates:update			✅

✳️ optional scope

Base Scopes

All base scopes must be granted in order to connect the integration with DataGrail. The remaining scopes are only required if enabling those capabilities

Endpoints Utilized

DataGrail uses the following endpoints to authorize and test the connection:

Method	Endpoint	Purpose	Docs
GET	/candidates	Validate the connection
POST	https://auth.greenhouse.io/token	Obtain and refresh an access token using client credentials

---

Limits

Limits in Greenhouse are calculated using the leaky bucket algorithm. All requests that are made after rate limits have been exceeded are throttled and an HTTP 429 Too Many Requests error is returned. Requests succeed again after enough requests have emptied out of the bucket.

• DataGrail supports requests throttling to stay within 70-80% of specified service rate limits.
• DataGrail processes API responses with HTTP 429 status to interrupt requests, waiting and retrying (using an exponential backoff strategy).

---

Capabilities

Access

DataGrail's Greenhouse integration provides Synchronous Access capabilities for the following supported identifier category: Email.

Data Interactions

For Access requests, DataGrail will take the following actions:

1. Search for Candidates using the Data Subject email address.
2. If a match is found, save the extracted candidate data in candidate_records, and then extract the candidate's applications to the separate apps_records collection.
3. Look up the jobs, rejection reasons, source, recruiter, and coordinator referenced by the applications and add them to each application, along with its current stage (built from the application's stage data). Recruiter and coordinator include only the ID, name, and employee ID, never the user's email address.
4. Retrieve the attachments for the candidate and their applications. Candidate attachments are added to the candidate record, and application attachments are added to each application. Each attachment includes its filename, type, creation date, and a signed URL that expires after 7 days.
5. Add the recruiter and coordinator to the candidate record, taken from the candidate's most recently updated application that has one.
6. Retrieve the candidates' offers and save them in the separate offers_records collection.

The jobs, rejection reasons, attachments, users (recruiters and coordinators), sources, and offers lookups are optional. If the credential doesn't have the required scope (Greenhouse responds with a 401 or 403), DataGrail skips that lookup, logs a warning, and completes the request with the remaining data. Any other error fails the request so it can be retried. Lookups for fields turned off in field-level controls are not made. Each request makes at most one batched call (up to 50 IDs) to the attachments, users, and sources endpoints.

Endpoints Utilized

Method	Endpoint	Purpose	Docs
GET	/applications	Retrieve a candidate's applications by candidate ID
GET	/attachments	Retrieve the attachments for a candidate and their applications by candidate ID
GET	/candidates	Retrieve a candidate by email
GET	/jobs	Retrieve the jobs referenced by a candidate's applications
GET	/offers	Retrieve a candidate's offers by candidate ID
GET	/rejection_reasons	Retrieve the rejection reasons (including type) referenced by a candidate's applications
GET	/sources	Retrieve the sources referenced by a candidate's applications
GET	/users	Retrieve the recruiters and coordinators referenced by a candidate's applications

---

Deletion

DataGrail's Greenhouse integration provides Synchronous Deletion capabilities for the following supported identifier category: Email. DataGrail supports two deletion modes: Deletion (Default) and Anonymization.

If you would like to change deletion modes, please contact support@datagrail.io.

Data Interactions

In Deletion mode, DataGrail will take the following actions for Deletion requests:

1. Delete a candidate by their ID.

In Anonymization mode, DataGrail will take the following actions for Deletion requests:

1. Update a candidate by their ID with the following anonymization rules:

Anonymization Rules

Set the following fields to null:

• addresses
• company
• custom_fields
• email_addresses
• phone_numbers
• social_media_addresses
• tags
• title
• website_addresses

Replace the following fields with a random string:

• first_name
• last_name

Application and offer data can also be anonymized when configured for your account:

• offers
• all_offer_versions
• custom_application_fields
• application_questions
• rejection_reason

Only the fields configured for your account are sent to Greenhouse.

Endpoints Utilized

Method	Endpoint	Purpose	Docs
PATCH	/candidates/ID	Anonymize a candidate
DEL	/candidates/ID	Delete a candidate

---