Zendesk AI agents — resolutionOutcome contract

Zendesk bills per automated resolution: the AI agent responded, the customer did not request further assistance, the conversation ended (channel-dependent window), and an LLM verification confirmed the request was resolved. Escalated conversations ('Assisted escalation') and conversations failing verification ('Contained') do not count.

Contract ID: zendesk_ai_automated_resolution
Source: https://support.zendesk.com/hc/en-us/articles/9570369117338-About-automated-resolution-tiers
Retrieved: 2026-09-11
Notes: Also: https://support.zendesk.com/hc/en-us/articles/5352026794010 and https://www.zendesk.com/pricing/ (retrieved 2026-09-11). Conversation end: email 72h after last email; messaging 2h default, configurable up to 72h; voice on hangup. The vendor's LLM verification step is internal and not observable from systems of record; the vendor lens encodes the structured parts only (no further assistance requested, no escalation) and is therefore at least as permissive as the vendor's own count. Per-resolution price and reopen handling are not published.

Vendor lens

Reproduces the vendor's published definition as closely as public documentation allows.

Verified if all of: no customer reply within 72h of the claim; escalated to a human within 72h of the claim
Not verified if any of: escalated to a human within 72h of the claim; a human agent replied within 72h of the claim; a duplicate claim exists for the same subject in the period
Unverifiable if any of: required fact source not connected: tickets

Strict lens

Applies the same definition with longer windows and additional checks a controller would expect.

Verified if all of: customer confirmed resolved within 30d of the claim
Not verified if any of: reopened within 30d of the claim; escalated to a human within 7d of the claim; a human agent replied within 7d of the claim; a refund or credit was issued within 30d of the claim; customer left a negative satisfaction rating; a duplicate claim exists for the same subject in the period
Unverifiable if any of: required fact source not connected: tickets; subject could not be matched in any fact source
Override (supporting evidence only): transcript label is customer_indicated_unresolved, abandoned → NOT_VERIFIED

Unknown fields

Things the source has not published. No value is assumed for these.

currency
llm_verification_criteria
reopen_handling
unit_price

Contract YAML

The machine-readable contract compiled into both lenses. Sign in to select it for your tenant or fork it.

# Zendesk AI agents — Automated resolution (Verified resolution tier).
id: zendesk_ai_automated_resolution
version: 1
vendor: zendesk
vendor_display: "Zendesk AI agents"
outcome_type: resolution
claim_source: zendesk.automated_resolutions
subject_kind: ticket
description: >
  Zendesk bills per automated resolution: the AI agent responded, the customer did not request further assistance,
  the conversation ended (channel-dependent window), and an LLM verification confirmed the request was resolved.
  Escalated conversations ('Assisted escalation') and conversations failing verification ('Contained') do not count.
source_citation:
  url: "https://support.zendesk.com/hc/en-us/articles/9570369117338-About-automated-resolution-tiers"
  retrieved: "2026-09-11"
  title: "About automated resolution tiers - Zendesk help"
  notes: >
    Also: https://support.zendesk.com/hc/en-us/articles/5352026794010 and https://www.zendesk.com/pricing/ (retrieved
    2026-09-11). Conversation end: email 72h after last email; messaging 2h default, configurable up to 72h; voice on
    hangup. The vendor's LLM verification step is internal and not observable from systems of record; the vendor lens
    encodes the structured parts only (no further assistance requested, no escalation) and is therefore at least as
    permissive as the vendor's own count. Per-resolution price and reopen handling are not published.
unknown_fields: [unit_price, currency, reopen_handling, llm_verification_criteria]
lenses:
  vendor:
    verified_if:
      all:
        - no_customer_reply_within: "72h"
        - not: {escalated_to_human_within: "72h"}
    not_verified_if:
      any:
        - escalated_to_human_within: "72h"
        - human_reply_within: "72h"
        - duplicate_claim_for_subject
    unverifiable_if:
      any:
        - missing_fact_source: [tickets]
  strict:
    verified_if:
      all:
        - customer_confirmed_resolved: {within: "30d"}
    not_verified_if:
      any:
        - reopened_within: "30d"
        - escalated_to_human_within: "7d"
        - human_reply_within: "7d"
        - refund_or_credit_within: "30d"
        - csat_negative
        - duplicate_claim_for_subject
    unverifiable_if:
      any:
        - missing_fact_source: [tickets]
        - identity_unmatched
    overrides:
      - when: {transcript_label_is: [customer_indicated_unresolved, abandoned]}
        verdict: NOT_VERIFIED
exclusions: []
billing:
  unit_price: null
  currency: USD
  unknown_fields: [unit_price]

Use this contract See it applied in a sample statement