pasochat

PASO is a full-stack real-time communication platform designed for scalability and intelligence. It integrates conventional chat functionalities with advanced features such as AI-powered moderation, voice and video calling, and horizontally scalable Socket.IO infrastructure. The platform supports 1:1 and group messaging, multi-instance communication via Redis, AI-driven smart replies, and ML-based analysis for toxicity, spam, and intent. It also includes file sharing, presence indicators, role-based administration, analytics, and automated CI validation. PASO was developed as part of ECSoC 2026 to demonstrate modern distributed web application architecture. Key challenges addressed include real-time communication with Socket.IO, multi-instance synchronization using Redis Pub/Sub, persistent state management with MongoDB, AI interactions via Groq API, automated moderation with a Python ML service, voice/video communication through ZegoCloud, and media storage with Cloudinary. The technology stack includes React for the frontend, Express.js for the API, Socket.IO for real-time communication, Redis for synchronization, MongoDB for persistence, and Python with FastAPI for the ML service. The architecture separates the real-time layer, REST API, AI/ML processing, persistence, and external integrations. The project provides detailed documentation for quick start, architecture, API, deployment, security, testing, and roadmap. It is suitable for development and production environments, with deployment options on platforms like Vercel and Render.