Call Forge Main Assistant

Workflow Overview Purpose & Automation Value This workflow automates the entire process of initiating an outbound, AI-powered phone call based on a simple natural language request. It acts as an intelligent "pre-call agent" that gathers necessary information, structures it, and then hands it off to a voice AI service to execute the call. The primary value is transforming a high-friction, manual task (looking up a contact, checking a calendar, dialing, and speaking) into a smooth, automated action. A user can simply state an objective like, "Call Sarah to schedule our project update for tomorrow," and the workflow handles the rest, saving significant time and ensuring a consistent, data-driven approach to outbound communication. Technologies Used - n8n: The core automation and orchestration platform. - LangChain: The framework used within n8n to build the intelligent AI agent. - Vapi.ai: The voice AI platform that executes the final phone call. - Anthropic (Claude 3.5 Sonnet): The Large Language Model (LLM) that powers the AI Agent's reasoning and text generation capabilities. - OpenAI (GPT-4o-mini): The LLM used by Vapi.ai to conduct the live phone conversation. - Airtable: Serves as the CRM/contact database. - Google Calendar: Used to check for scheduling availability. - PostgreSQL: Functions as a memory store for retrieving past conversation history. AI Integration Analysis This workflow showcases a sophisticated, dual-layer AI architecture: 1. Pre-Call Preparation AI (n8n & LangChain) The core of the n8n workflow is the AI Agent node. This is not the AI that speaks on the phone; it's the AI that _prepares_ for the phone call. Its function is to: - Understand Intent: It uses the Anthropic Claude 3.5 Sonnet model to parse the user's natural language query (e.g., "Call John..."). - Gather Context with Tools: The agent is empowered with tools to autonomously fetch required data. It can query Airtable for a contact's phone number, check the user's Google Calendar for open time slots, and access conversation history from a PostgreSQL database.