WAgent Chatbot

Multi-Platform Autonomous AI Assistant for WhatsApp & Messenger

Category: ai-automation | Role: AI Automation Developer — Architected multi-node agent workflows, connected messaging webhooks, integrated productivity tool APIs, and tuned agent reasoning prompts.

Project Overview

WAgent Chatbot is an autonomous multi-platform agent engineered across two major production workflows: WAgent_WhatsApp_Evolution_API (58 nodes) and WAgent_Messenger_Voice_Multiplatform (55 nodes). The assistant handles text and audio interactions across WhatsApp (via Evolution API) and Facebook Messenger (via Meta API). Audio messages are transcribed on-the-fly using OpenAI Whisper. An autonomous GPT-4 agent utilizes Buffer Window Memory for conversational context and leverages dynamic tool calling to search the web (Perplexity AI, SerpAPI, Wikipedia, HackerNews), manage scheduling (Google Calendar), read and send emails (Gmail), append data to Google Sheets, and publish updates to Twitter/X.

Key Engineering Highlights

  • Two Production Workflows: Built WhatsApp Evolution API flow (58 nodes) and Messenger flow (55 nodes)
  • Multi-Platform Messaging: Unified messaging integration connecting Evolution API and Meta Webhooks
  • Voice Message Processing: Transcribes incoming voice notes into actionable prompt contexts with OpenAI Whisper
  • Autonomous Tool Calling: Executes real-time web research across Perplexity AI, SerpAPI, Wikipedia, and HackerNews
  • Productivity Integrations: Automates Google Calendar scheduling, Gmail messaging, and Google Sheets logging
  • Conversation Memory: Maintains contextual dialogue history across multi-turn interactions using Buffer Window Memory

Technology Stack

n8nOpenAI GPT-4OpenAI WhisperWhatsAppFacebook MessengerEvolution APIMeta APIPerplexity AISerpAPIGoogle CalendarGmailGoogle SheetsTwitter/XWikipediaHackerNewsBuffer Window Memory