🏢 Business / Industry: Sales / B2B outbound

🎯 Main Goal: Automatically generate personalized call scripts for each lead based on CRM data and deliver them to sales managers via Telegram in under 10 seconds.

🔹 Context

Sales teams with active pipelines rely on static, one-size-fits-all call scripts. Managers spend 5–10 minutes before each call manually reviewing CRM notes, recalling previous conversations, and preparing an opening. Across a full day, this adds up to 1–2 hours of unproductive preparation — and the resulting scripts are still generic and impersonal.

🔹 Problem

The sales department used a static 10-page document as their primary call guide. It outlined the general logic of client interaction but had critical limitations: managers read it once and rarely revisited it; it was identical for every lead regardless of funnel stage; it became outdated within a month; and it provided no context about previous communication or recent changes since the last contact. Average conversion with this approach was 15–20%.

🔹 Solution

An n8n workflow was built to replace static scripts with dynamic, AI-generated call scenarios. The manager sends a Lead ID to a Telegram bot, the workflow pulls the full lead profile from Zoho CRM, and an LLM generates a personalized script tailored to the lead's funnel stage, communication history, and recent context changes. The ready-to-use script is delivered back in Telegram within 10 seconds — before the call starts.

🔹 Process

• Telegram Trigger: The workflow activates when a sales manager sends a Lead ID to the Telegram bot, initiating the script generation pipeline.
• Lead ID parsing (Code): A Code node extracts and validates the Lead ID from the incoming message, preparing it for the Zoho CRM API request.
• Zoho CRM: Get Lead (HTTP Request): An HTTP Request node queries Zoho CRM using the extracted Lead ID, retrieving the full lead profile — including funnel stage, previous communication, last interaction date, and CRM notes.
• AI: Script generation (Basic LLM Chain): The enriched lead data is passed to a Basic LLM Chain. The model determines the call type based on the funnel stage and generates a structured, personalized call script that references specific details from the CRM record.
• Telegram response: The generated script is immediately sent back to the manager via Telegram — formatted, contextualized, and ready to use before the call.