Use Case Intelligence RSS Search use cases Search
Anthropic Software / SaaS North America

Zendesk

Customer Service Automation

Zendesk wanted customer-specific service agents to handle the complex work between a request and its resolution, but one-off automations were expensive to build, hard to localize, and difficult to govern across millions of service interactions.

Read the original story Explore all use cases
1 million custom agent executions in the first seven weeks of early access
99% accuracy in its pilot for duplicate merging
Raised automated resolution by up to 10%
Up to 80% reduction in human handle times

Reported impact

Business outcome

Zendesk reached 1 million custom-agent executions in the first seven weeks of early access, helped pilot customers cut human handle times by up to 80 percent, raised automated resolution by up to 10 percent, and proved that custom agents can be built with a small central team instead of one-off AI projects.

  • Compliance
  • Customer Experience
  • Productivity Improvement

Solution

Zendesk built its custom-agent builder on the Claude Platform with a five-person team using Claude Code and the Claude Agent SDK. Customers describe the work in plain English and connect the agent to Zendesk actions plus external tools such as Jira or Google Drive, turning workflows that once needed AI specialists into a roughly 30-minute setup. Zendesk evaluated models on 15 metrics using realistic customer-service agents and selected Claude Sonnet 4.6 because it read PDFs well, reached conclusions in fewer planning turns, and delivered better cost per task. Production runs through Amazon Bedrock in Zendesk's existing AWS regions so data locality and security controls stay inside the environment the company already governs. That architecture lets Zendesk deliver specialist agents for returns, warranty questions, billing exceptions, and other business-specific service work.

Data flow

A Zendesk customer writes instructions in plain English, connects internal Zendesk actions and external tools such as Jira or Google Drive, and the agent then handles the service workflow on Claude Sonnet 4.6 through Amazon Bedrock while humans keep final operational oversight.

Solution architecture

5 components · 4 layers
  1. Compute
    • Claude Code Helped the five-person team build the custom-agent builder from proof of concept to early access.
    • Claude Sonnet 4.6 Runs customer agents that need strong document reading and low planning-turn overhead.
  2. Serving
    • Claude Platform Provides the base model platform for Zendesk's custom service agents.
  3. Orchestration
    • Claude Agent SDK Supplies the agent-building layer behind Zendesk's configurable service workflows.
  4. Uncategorized
    • Amazon Bedrock Keeps Claude running inside Zendesk's existing AWS regions for security and data locality.

Architecture clues

  • Claude runs through Amazon Bedrock inside the AWS environment Zendesk already governs.
  • Customers can connect agents to Zendesk actions and external tools in about 30 minutes.
  • Zendesk scores candidate models on 15 metrics using agents like the ones customers actually write.

Evidence from the source

Built its custom agent builder with a five-person team, working with Claude Code and the Claude Agent SDK
It takes about 30 minutes to set up something that used to take
Raised automated resolution by up to 10% while cutting human handle times by up to 80% for customers running custom agents
Went from zero to 1 million custom agent executions within seven weeks of early access
Zendesk chose Claude Sonnet 4.6.
Transparency Evidence & methodology 100/100 reference value · 79% confidence · 1 source

Review how this reference was scored and trace each normalized claim back to its documented source.

01

Scoring methodology

How scores are calculated

Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.

Customer reference value100/100
  • Problem clarity20/20
    Measures whether the customer's operational problem is specific enough to compare.
  • Implementation depth25/25
    Rewards named components, data flow, integrations, and architecture detail.
  • Outcome specificity20/20
    Measures whether the reported benefit is concrete rather than a generic claim.
  • Quantified impact25/25
    4 metrics describe measurable impact.
  • Deployment scale10/10
    Rewards evidence of a live or scaled deployment and meaningful integration scope.
Evidence confidence79%
  • Source authority20/25
    The main evidence is an official vendor-published customer story.
  • Claim traceability21/25
    10 of 12 key fields link to a named source.
  • Direct evidence20/20
    5 direct source quotes retained.
  • Date provenance0/10
    The publication date has no documented provenance.
  • Source transparency8/10
    1 source reference document publisher, URL, authority, retrieval, and supported fields.
  • Implementation specificity10/10
    Rewards source-backed components, data flow, quantified impact, and a specific outcome.
02

Traceability

Sources & provenance

Field coverage shows which facts each source supports. Dates are shown only when documented by the source.

  1. Anthropic

    Zendesk customer story

    First-party
    https://claude.com/customers/zendesk
    Source type
    Vendor-published customer story
    Supports
    • Customer
    • Business problem
    • Solution
    • Products
    • Solution components
    • Data flow
    • Integrations
    • technical area
    • use case category
    • Business outcome
    • outcome category
    • Metrics
    • Architecture
    • evidence quotes

Date evidence

  • First archived Oct 8, 2026 · Internet Archive · Independent First Wayback Machine capture; confirms the page existed by this date, not when it was published.
    Open date source
Not separately sourced Industry, Region, Company size. Treat these normalized attributes as lower-confidence until a dedicated source is added.

Related

Similar use cases

See all like this
Anthropic Aug 26, 2026 Rocket Money Software / SaaS North America Rocket Money wanted Rowan to deliver financial help by text message without making users open dashboards, while still keeping the assistant accurate enough to act on real money decisions. Customer Service Automation Claude Platform Anthropic Aug 26, 2026 Rocket Money Software / SaaS North America Rocket Money wanted to put a trustworthy financial expert in every customer's pocket, but a personal finance agent that reads account data, answers questions, and takes actions has to be highly accurate, observable, and… 11x increase in code commits from 11 in December to 128 in July Customer Service Automation Amazon Bedrock · Claude Code · Claude Platform Google Cloud Modified · Oct 9, 2026 Tally Group Software / SaaS Australia This rapid processing surfaces targeted insights in under 20 seconds, eliminating manual checklists. At utility billing platform Tally Group , call center agents previously navigated up to six applications to… Customer Service Automation Apigee API Management · BigQuery · Gemini platform