Back to docs
API Reference

Zendesk + Koji: Trigger AI Interviews from Support Tickets and Sync Insights Back to Every Conversation

Use Zendesk triggers to invite customers to Koji AI-moderated interviews from solved tickets, low CSAT, or specific tags, tie each interview to its ticket through the headless API, and have a small scheduled worker post themes, sentiment, and a response link back onto the ticket.

Answer first: A Zendesk ticket tells you a customer had a problem; it rarely tells you why the problem mattered, what they expected, or whether they are now a flight risk. Koji closes that gap. The integration works in both directions: (1) a Zendesk trigger invites the customer to a Koji AI-moderated interview when a ticket is solved, a CSAT rating comes in low, or a specific tag is applied, and (2) once the interview is analyzed, a small worker posts a concise insight summary back onto the ticket as an internal note and tags: sentiment, top themes, quality score, and a link to the response. No native Zendesk app is required. Zendesk's own triggers handle the invitation, Koji's headless API ties an interview to a ticket ID, and a Koji webhook handles the way back: each time a customer finishes, Koji sends your small service a signed message carrying the ticket ID you passed in. A first version is live in well under an hour.

The result is a support org that does not just resolve tickets. It systematically learns from them, at a scale no manual QA or callback program can match.

Why connect Zendesk and Koji

Support conversations are the largest, most honest stream of customer signal most companies own, and the least mined. The ticket text captures the symptom ("export is broken") but almost never the stakes ("I nearly cancelled because I present this to my CEO every Monday"). Traditional follow-up options are weak: a one-question CSAT survey with no depth, or an agent-initiated callback that does not scale past a handful a week.

Koji's AI moderator interviews the customer conversationally, asks adaptive follow-up questions, and works in text or voice across 30+ languages, so you can follow up on hundreds of tickets a week without adding headcount. Every resolved ticket can become a two-minute "what really happened" interview, and every low CSAT score can trigger a structured root-cause conversation instead of a lost customer.

What flows in each direction

Zendesk → Koji (invite customers from ticket events)

  • A Zendesk trigger fires on an event you choose (ticket solved, CSAT = bad, tag churn-risk added, priority = urgent escalated).
  • Simplest version: the trigger's email notification to the requester includes your study's interview link. Koji runs and analyzes the interview, but it does not know which ticket it came from, so this suits aggregate learning rather than ticket-level write-back.
  • Ticket-linked version: the trigger's webhook calls your own small service, which calls Koji's Start Interview endpoint with a respondent object: the ticket ID as external_id, the requester's name as display_name, and anything else useful (ticket tags, locale) in metadata. The endpoint returns an interview_id and a session token, and your service runs the conversation in your own help-center page or widget through the message endpoint. The study is resolved from the API key, so each key belongs to one study.
  • On the ticket-linked path, Zendesk or your own mailer delivers the invitation. For batch follow-ups, Koji can also send the invitation email itself (see path 2).

Koji → Zendesk (insights land back on the ticket)

  • Add a webhook to the study (Integrate, then Webhooks). Each time a customer finishes, Koji POSTs a signed interview.completed message to your service, retried if it is down and logged so you can resend. Webhooks are on every paid plan, not Free.
  • The message carries the external_id you set at start (your ticket ID) and a quality score, so the service knows which ticket to update. It then calls GET /api/v1/interviews/{id} for the themes, sentiment and summary. On Free, a scheduled worker polling that endpoint does the same job.
  • The worker calls the Zendesk API to add an internal note to the original ticket with the top themes, sentiment, quality score, and a link to the response in Koji, and to apply tags like insight:pricing-confusion or sentiment:frustrated.
  • Agents and CX leads see the "why" attached to the very ticket that prompted it.

Ticket events worth interviewing

  • Solved tickets: a lightweight "did we actually fix the underlying job?" check.
  • Low CSAT / detractor ratings: a structured root-cause interview while it is fresh (see churn interview questions).
  • Churn-risk or cancellation tags: route straight to a cancel-flow exit interview.
  • Feature-request tickets: probe the underlying need, not just the asked-for feature.
  • Repeated escalations: understand the cumulative frustration a single ticket hides.

Setup: three paths

1. Zendesk trigger with your study link (no code). Create a trigger whose conditions match the event and whose action emails the requester with your Koji interview link. You get themed, aggregated insight in the Koji report within minutes of setup. There is no per-ticket write-back on this path.

2. Personalised invite links in batches. Export the requesters you want to hear from (for example, last week's detractors) and import them into Koji as a CSV. Koji creates a personal interview link for each person, and each response is tied to that person in Koji. Send the links from Zendesk or your mailer, or import the requesters into Participants, save them as an audience and let Koji email the invitation and a reminder from the Self-recruit tab of the study. Each person gets their own link, and you can see who received the email, opened it and started the interview. See Connect Koji to Zapier if you want to automate the export.

3. Headless API from your own middleware (ticket-linked). If you already run a support-automation service, have the Zendesk webhook call it, start interviews through Koji's headless API with the ticket ID as external_id, and let a Koji webhook tell it when each interview is done; see Webhook Setup for the payload and the signature check.

Design the interview so it maps to your taxonomy

Koji's structured questions make the returned data align with how your CX team already thinks. A study can mix six types: open_ended, scale, single_choice, multiple_choice, ranking, and yes_no. A single_choice "which best describes the root cause?" maps directly to your ticket-reason taxonomy, while open_ended probes capture the story. See the structured questions guide. Every quantitative answer aggregates into a real-time report you can watch fill in as interviews complete.

Example: turning CSAT detractors into a save motion

  1. Zendesk trigger: CSAT rating = bad → webhook to your service, which starts a Koji interview with the ticket ID as external_id.
  2. The customer gets a 3-minute interview: one single_choice root-cause question, two open_ended probes, one scale "how likely are you to keep using us?"
  3. When the webhook arrives, your service reads the analysis, tags the ticket sentiment:at-risk, posts the summary as an internal note, and alerts the account owner in Slack.
  4. CX now has a ranked, themed view of why CSAT dips, and a triggered save conversation for every detractor.

Identity and privacy

On the headless path, pass the Zendesk ticket ID (or user ID) as respondent.external_id. The same value comes back when you read the interview, so notes land on the right ticket. Keep email addresses out of metadata unless you need them. For sensitive topics, use the plain study link instead: Koji does not know who the person is, and you skip the Zendesk write-back, keeping raw transcripts out of the helpdesk.

Plans and limits

The headless API, API keys, the MCP connector, CSV import with personalised links and voice interviews work on every account, Free included: credits are the gate. Sharing a study link from a Zendesk trigger works on every plan too. Recruitment emails sent by Koji and webhooks come with every paid plan, not Free, and recruitment emails have a monthly allowance with extra emails available when you need more. Credits cost €1 each: a qualified text interview uses 1 credit and a qualified voice interview uses 3.

The insight note that lands on the ticket

When an interview is analyzed and your worker writes it back, the internal note it posts on the ticket is designed to be skimmable by a busy agent. A typical note includes the overall sentiment, the two or three ranked themes, the answer to each structured question, the interview quality score, and a link to the full response. Tags such as insight:onboarding-gap and sentiment:at-risk are applied to the ticket, so you can build Zendesk views and reports around them, for example a view of every "at-risk" customer surfaced by a follow-up interview this week.

Common pitfalls (and how to avoid them)

  • Triggering on every ticket. Volume kills goodwill. Gate interviews to the moments that carry real signal (bad CSAT, cancellation tags, complex solved issues) and cap frequency per requester.
  • Asking what the ticket already answered. Do not re-ask the reported issue. Use the interview to probe stakes, expectations, and alternatives considered, which the ticket never captures.
  • Letting insights die in a note. Route the highest-severity findings (negative sentiment plus a high quality score) to Slack or a weekly CX review so they drive action, not just documentation.
  • Ignoring language. Koji moderates in 30+ languages; pass the ticket's locale when you start an interview so international customers get a native-language conversation.

From tickets to a continuous voice-of-customer program

Handled this way, Zendesk stops being only a cost center and becomes your largest research panel. Every week, solved tickets and CSAT detractors feed a themed, quantified stream of root-cause insight, the kind of signal product and CS leaders usually pay agencies for. Because Koji analyzes each interview automatically and aggregates structured answers into a live report, the CX team gets a standing "why customers contact us" view without running a single manual study, and on the ticket-linked path every new theme is traceable back to the exact ticket and response that produced it.

Related Resources

Related Articles

API Authentication

How Koji Headless API authentication works: pk_live_ API keys, permissions, session tokens, and origin allowlists.

Cancel-Flow Exit Interviews: AI Moderation That Saves Churning Users

Replace your single-textarea cancellation survey with an AI-moderated exit interview that probes the real reason people leave and triggers save offers in real time.

Churn Interviews: 20 Questions to Uncover Why Customers Really Leave (2026)

A complete guide to running customer churn interviews: when to interview vs survey, who to talk to, 20 non-leading questions grouped by the push-pull framework, and how to automate churn interviews with AI on Koji.

How to Use Your CRM Data for Targeted AI Research: Import Participants and Personalize Every Interview

Your CRM already contains your best research sample. Learn how to export customer segments, import them into Koji, and invite each person with a personal link, either by Koji email with reminders and tracking or from your own CRM.

Customer Success Interview Guide: How to Run Proactive CS Research at Scale

How customer success teams use AI-powered interviews to proactively understand account health, expansion opportunities, and churn risk — without scheduling 1:1 calls with every customer.

Free Trial Feedback: How to Learn Why Trials Don't Convert (2026)

A practical guide to free-trial feedback for SaaS teams: what to ask trial users who don't convert, how to interview them automatically, and how Koji's AI surfaces the activation gaps that quietly kill your conversion rate.

Headless API Overview

Manage interviews programmatically with the Koji REST API — start, message, and complete interviews from your own code.

Sync Koji Customer Interviews to HubSpot: Live Insights on Every Contact

Push Koji interview transcripts, themes, and quality scores onto HubSpot contact and company records using the Koji API and the HubSpot API.

Intercom + Koji: Run Deep AI Interviews From Your Support and Messaging Workflows

Trigger Koji AI-moderated interviews from Intercom conversations, post-resolution surveys, and outbound messages — then sync themes, sentiment, and quality scores back as Intercom user attributes.

Real-Time Research Insights: How to See Themes, Quotes, and Quality Scores as Interviews Complete

Stop waiting weeks for analysis — modern AI research platforms surface themes, structured-question distributions, sentiment, and quality-scored quotes the moment each interview ends. Here is how real-time research insights work in Koji and how to design studies that take advantage of them.

Research Automation: How to Build Real-Time Research Pipelines with Webhooks

Build automated research pipelines on Koji: event-triggered interviews, a signed webhook when each one is finished and scored, and insight routing to Slack, your CRM or a warehouse.

Sync Koji AI Interviews to Salesforce: Customer Insights on Every Account, Contact, and Opportunity

Push Koji interview quotes, themes, sentiment and quality scores into Salesforce with a webhook-driven worker, or send personal interview links from a Salesforce report with no code.

Send Research Insights to Slack: Real-Time Customer Interview Notifications via Webhooks

Get customer interview insights from Koji into Slack. There is no native Slack app, but Koji sends a signed interview.completed webhook from any study (Integrate, then Webhooks), which a small receiver or Zapier, Make or n8n turns into a Slack message seconds after each interview. Step-by-step setup with message recipes.

Starting Interviews via API

Create interview sessions programmatically with POST /interviews/start: request fields, the 201 response, CRM linking, and voice mode.

Structured Questions in AI Interviews

Mix quantitative data collection — scales, ratings, multiple choice, ranking — with AI-powered conversational follow-up in a single interview.

Understanding Themes & Patterns

Learn how Koji identifies recurring themes across interviews and how to use them for decision-making.

User Research API: Embed AI Interviews into Any Product or Workflow

How to use Koji's User Research API to run AI-moderated interviews from your own backend. Covers REST endpoints, the embed widget, signed completion webhooks, authentication, rate limits, and headless interview patterns.

Webhook Setup

Set up Koji webhooks: get a signed interview.completed message on your server each time a participant finishes, with your own participant ID and a quality score.

Connect Koji to Zapier: Automate Customer Research Workflows in Minutes

Route every completed AI customer interview from Koji into 6,000+ Zapier apps — including Notion, Linear, Salesforce, Airtable, and Gmail. A step-by-step integration guide.

Segment + Koji: Trigger AI Interviews from CDP Audiences and Sync Insights Back as Traits

Use Segment Audiences and events to trigger Koji AI-moderated interviews, then pipe summaries, key insights and sentiment back into Segment as track events and user traits that fan out to every downstream tool.