Zapier

With the Zapier integration, you can take RepX conversation data and push it to any product in your enterprise tech stack.

RepX fires an event every time a conversation completes, carrying the entire conversation and everything RepX extracted from it - the summary, topics, transcript, lead details, and all of your conversation variables.

• Setting up the Zapier integration

• What the event contains

• Use cases: What teams build with it

Setting up the Zapier integration

• Go to Settings > Integrations from the left navigation panel on RepX dashboard.

• Choose the Zapier integration and click Connect.

• You'll be given an API key. Copy it.

• In Zapier, create a new Zap, choose Storylane as the trigger app, and paste the API key when Zapier asks you to connect your Storylane account.

• Select RepX conversation completed as the trigger event.

Zapier will pull data from a sample conversation so you can get a sense of the attributes we send over.

The RepX conversation completed event

The repx_conversation_completed event fires once for every conversation, as soon as that conversation is complete and RepX has finished processing it. A conversation is considered completed when the user closes the RepX chat panel or when the user closes the browser tab where they were having a RepX conversation.

What the event contains

Field	Data Type	Description
id	String	Unique identifier for the conversation
event	String	Always repx_conversation_completed
transcript	String	The full conversation, message by message, including whether it happened in text or voice mode
conversation_summary	String	A short AI-generated summary of what the conversation covered
topics	Array	The topics discussed during the conversation
intent_level	String	Buying intent calculated for the conversation - high, medium or low
extracted_variables	Object	All conversation signals extracted from the conversation - both default and custom
duration	Number	Length of the conversation, in seconds
message_count	Number	Number of messages exchanged in the conversation
started_at	String	Timestamp of when the conversation began, in ISO 8601
completed_at	String	Timestamp of when the conversation completed, in ISO 8601
cta_opened	String	The CTA URL the visitor clicked during the conversation, if any
lead	Object	Lead details - id, email, first_name, last_name, lead_source, client_source and client_tracking_id
buyer_reveal	Object	Company details resolved from the visitor's IP or email - name, domain, revenue range, employee range and a confidence score
location	Object	The visitor's country, city and state
conversation_dashboard_url	String	Contains the link to the conversation in RepX dashboard

extracted_variables always contains the four default signals - email, visitor_name, phone_number and qualified. See Conversation signals for how to set them up.

Sample payload:

{
  "id": "2f37f8ce-6ecb-49a2-be12-36ae9c063042",
  "event": "repx_conversation_completed",
  "buyer_reveal": {
    "company_name": "Acme Inc",
    "company_domain": "example.com",
    "company_annual_revenue_range": "1M-10M",
    "company_employee_range": "11-50",
    "confidence_score": "high"
  },
  "lead": {
    "id": "f3694255-aed4-4514-8ce6-ac55506a67fe",
    "client_source": "none",
    "client_tracking_id": null,
    "email": "[e-mail]",
    "first_name": "John",
    "last_name": "Doe",
    "lead_source": "Direct"
  },
  "intent_level": "high",
  "conversation_summary": "The user asked about pricing and integrations while planning a migration, and the assistant shared a calendar link to book a live demo.",
  "topics": ["Pricing", "Integrations", "Customer Onboarding"],
  "duration": 90,
  "messages_count": 3,
  "extracted_variables": {
    "email": "[e-mail]",
    "visitor_name": "John Doe",
    "phone_number": "[phone]",
    "qualified": true
  },
  "transcript": "[text mode]\n\nAssistant: Hi, how can I help you today?\n\nUser: I want to see a demo.\n\nAssistant: Are you a new customer?\n\nUser: Yes, we're planning a migration soon.\n\nAssistant: Got it. What's your work email?\n\nUser: [e-mail]\n\nAssistant: Thanks, John. Pick a time on the calendar to book a live demo.",
  "started_at": "2026-07-22T10:54:05.337+00:00",
  "completed_at": "2026-07-22T10:55:35.337+00:00",
  "cta_opened": "https://example.com",
  "location": {
    "country": "United States",
    "city": "Miami",
    "state": "FL"
  }
}

Usecases with Zapier

Here are some usecases customers often build with the Zapier integration:

• Brief a rep before they reach out. When a visitor submits their email, send the summary, topics and pain points to the BDR who owns that account - in Slack, by email, or as a task in your sales engagement tool. The rep opens the conversation already knowing what the prospect cares about instead of starting cold.

• Push custom signals into specific CRM fields. The HubSpot integration writes a standard set of conversation attributes as properties and notes. Anything beyond that - pain points, number of users or sites, primary product interest - can be mapped through Zapier into whichever custom fields your CRM already uses, so the information sits where your team already looks for it. Because signals arrive typed, a Number signal lands in a numeric CRM field you can build workflows against.

• Send the transcript to the visitor. Take the email and transcript from the event and trigger an automated email through your nurture tool as soon as the conversation ends, so the visitor has a record of what they were told.

• Post high-intent conversations to Slack. Filter on intent_level or on qualified being true, and only notify your team about the conversations worth acting on today.

• Route by product interest. Use a custom signal to branch the Zap - conversations about one product go to one team's channel or queue, conversations about another go elsewhere.

• Score and enrich before handoff. Send the conversation into your enrichment or scoring tool first, then create the CRM record only once it clears your bar, so your CRM stays clean.

Your endpoint should return a 2xx response as soon as it receives the event, and do any heavy processing afterwards. Acknowledge first, process second.