Conversation Events

Overview

Asiyst provides a comprehensive event emitter on the client SDK and webhook notifications on the server to notify your application whenever a visitor starts a chat, sends a message, receives an answer, or leaves rating feedback.

Client SDK Events

Subscribe to lifecycle events directly in your frontend code:

import { Asiyst } from "@asiyst/sdk";
// Conversation opened
Asiyst.on("conversation:started", ({ conversationId }) => {
  console.log("Started conversation:", conversationId);
});
// User dispatched a query
Asiyst.on("message:sent", ({ text, timestamp }) => {
  console.log("User asked:", text);
});
// Assistant completed generation
Asiyst.on("message:received", ({ text, citations, latencyMs }) => {
  console.log("Assistant responded in", latencyMs, "ms");
});
// Visitor submitted feedback thumbs up / down
Asiyst.on("feedback:submitted", ({ messageId, rating, comment }) => {
  console.log("Feedback:", rating, comment);
});

Server Webhook Events

Configure HTTP endpoints via the Asiyst Webhooks console (/dashboard/api/webhooks) to receive server-side notifications:

Webhook Payload Example:

{
  "event": "conversation.completed",
  "timestamp": "2026-09-28T03:30:00Z",
  "data": {
    "conversationId": "conv_9a8b7c6d5e",
    "projectId": "proj_9f8a2b3c4d",
    "turnCount": 4,
    "durationSeconds": 142,
    "satisfactionScore": 5
  }
}

Capturing User Feedback & Ratings

When visitors click the thumbs-up or thumbs-down buttons beneath assistant answers:

• The rating is stored alongside the message transcript.

• Feedback is aggregated in the Analytics > Conversation Quality dashboard.

• If negative feedback is submitted with a comment, an alert can be dispatched to your team's Slack or Discord channel.

Integrating with External Analytics

Forward Asiyst conversation events to Google Analytics, PostHog, or Segment:

Asiyst.on("conversation:completed", (data) => {
  // Google Analytics 4
  if (typeof gtag === "function") {
    gtag("event", "assistant_conversation_completed", {
      conversation_id: data.conversationId,
      turn_count: data.turnCount,
    });
  }
  // PostHog
  if (window.posthog) {
    posthog.capture("Assistant Conversation Completed", data);
  }
});