Web Automation

Snapchat Web Messaging Automation System

A desktop-based automation system for managing Snapchat Web conversations, supporting both manual messaging and AI-powered automated replies.

Project Overview

This project is a web-based messaging automation system built on top of Snapchat Web. It allows users to send messages manually to specific usernames or automatically manage incoming conversations at scale. All message activity is persisted for tracking, analysis, and follow-ups.

Purpose & Scope

The purpose of this system is to automate Snapchat Web messaging workflows while keeping conversations observable and manageable. Its scope includes manual outbound messaging, automated inbox scanning, AI-assisted replies, conversation storage, and desktop-based orchestration.

Core Features

  • Manual message sending to specific Snapchat usernames
  • Automated scanning of full DM conversation lists
  • Detection of latest incoming messages
  • AI-generated replies using OpenAI
  • Conversation history persistence
  • Desktop-level automation via Electron

Automation Workflow

The workflow supports both manual and automated modes. In manual mode, the system opens Snapchat Web, navigates to a target user, sends the requested message, and stores the result. In automated mode, it scans the inbox, detects the latest inbound messages, saves chat context, sends it to the AI layer, and posts the generated reply automatically.

System Architecture

The desktop application serves as the orchestration layer, managing browser sessions and automation logic. Puppeteer handles interaction with Snapchat Web, while the Python backend coordinates workflows, AI responses, and database operations. This separation allows for scalable messaging automation without relying on unofficial APIs.

Technology Stack

Python Electron.js Puppeteer Snapchat Web OpenAI APIs Database Persistence

Use Cases

Built for inbox automation, support-style response workflows, conversation monitoring, and AI-assisted messaging operations where manual outreach and automatic replies need to coexist in one controlled system.