Zendesk

API

The Zendesk connector is an API connector for Zendesk Support. It imports tickets using cursor-based pagination and produces structural claims for status, priority, tags, and title. Ticket descriptions are run through text extraction for additional claims.

How it works

Queries the Zendesk Tickets API (/api/v2/tickets) with API token or OAuth
Uses cursor-based pagination for efficient traversal of large ticket volumes
Produces structural claims per ticket: status, priority, tags, and title
Ticket descriptions are optionally run through text extraction
Supports server-side status filtering (e.g. only open or pending tickets)
Authentication via API token (email + token) or OAuth Bearer token

Claim patterns

Subject	Predicate	Object
`(ticket id, ticket)`	`has_status`	`(status, status)`
`(ticket id, ticket)`	`has_priority`	`(priority, level)`
`(ticket id, ticket)`	`tagged`	`(tag, tag)`
`(ticket id, ticket)`	`titled`	`(subject, title)`

1 Create a Zendesk API token

In your Zendesk Admin Center, go to Apps and integrations > APIs > Zendesk API and create an API token. You will also need the email address associated with the token.

2 Install requests

$ pip install requests

3 Connect

conn = db.connect("zendesk",
    subdomain="mycompany",
    email="admin@mycompany.com",
    token=os.environ["ZENDESK_API_TOKEN"],
)
result = conn.run(db)

Parameter	Required	Default	Description
`subdomain`	Yes	—	Zendesk subdomain (e.g. `mycompany` for mycompany.zendesk.com)
`token`	Yes	—	Zendesk API token or OAuth Bearer token
`email`	No	—	Email for API token auth (used as `{email}/token`)
`status`	No	All	Filter tickets by status (e.g. `open`, `pending`, `solved`)
`max_tickets`	No	`500`	Maximum number of tickets to fetch
`extraction`	No	`heuristic`	Text extraction mode for descriptions: `heuristic` or `llm`
`save`	No	`False`	Encrypt and persist credentials

Basic usage

conn = db.connect("zendesk",
    subdomain="mycompany",
    email="admin@mycompany.com",
    token=os.environ["ZENDESK_API_TOKEN"],
)
result = conn.run(db)
print(f"Ingested {result.claims_ingested} claims from {result.tickets_processed} tickets")

Filtered by status

conn = db.connect("zendesk",
    subdomain="mycompany",
    email="admin@mycompany.com",
    token=os.environ["ZENDESK_API_TOKEN"],
    status="open",
    max_tickets=200,
)
result = conn.run(db)

With LLM extraction

conn = db.connect("zendesk",
    subdomain="mycompany",
    token=os.environ["ZENDESK_OAUTH_TOKEN"],
    extraction="llm",
    save=True,
)
result = conn.run(db)