DisChat – High-Performance Real-Time Chat Platform
A Discord-inspired real-time chat platform built with a blazing-fast Rust backend, WebSocket connections, dual-database architecture (PostgreSQL + MongoDB), and a modern TypeScript frontend.
Vue d'ensemble du projet
DisChat is a high-performance, Discord-like chat platform engineered for scalability and real-time responsiveness. The backend is written in Rust using the Actix-Web framework, delivering sub-millisecond message latency and effortless handling of thousands of concurrent WebSocket connections.
The architecture uses a dual-database strategy: PostgreSQL manages relational data such as users, channels, and permissions, while MongoDB stores chat message history for flexible, high-throughput document storage. This combination ensures both data integrity and write performance at scale.
The frontend is built with TypeScript, delivering a responsive and type-safe user experience. Features include server/channel organization, direct messages, real-time presence indicators, file sharing, and role-based access control — mirroring the core workflow of Discord with a lean, modern stack.
Fonctionnalités clés
Technologies utilisées
Détails du projet
Client
Projet Personnel
Timeline
2025 – Présent
Rôle
Architecte & Développeur Full-Stack
Autres projets

JobHub – Plateforme d'Emploi Intelligente avec Agent IA
Application Web Full-Stack

Email AI Platform
Intelligence Artificielle

Ovia
Intelligence Artificielle

MediBook Platform
Web Application (Healthcare SaaS)

MediBook Healthcare Passport
Web Application (Healthcare SaaS)

MediBook Pharma Platform
Web Application (Healthcare SaaS)

MediBook Laboratory Platform
Web Application (Healthcare SaaS)

Personal Portfolio Website
Web Application
Reboturn
E-Commerce Platform (Eco-Conscious SaaS)

Salary Prediction with Linear Regression
Machine Learning Project

Insurance Cost Prediction with Multivariable Linear Regression
Machine Learning Project

Logistic Regression (Classification)
Machine Learning Project

Brain Tumor Classification (Deep Learning - CNN)
Machine Learning Project

Decision Trees
Machine Learning Project

Random Forest
Machine Learning Project

Clustering (K-means)
Machine Learning Project

Anomaly Detection
Machine Learning Project

Système de Recommandation Netflix
Intelligence Artificielle

Recommender Systems (Content-based Filtering)
Machine Learning Project

VelociType — Test de Vitesse de Frappe
Application Python

Watermark — Application de Filigrane d'Images
Application Python

XO Battle — Jeu de Tic-Tac-Toe
Jeu Python

PulseCode — Morse Converter
Application Python
© 2026 Sina Ganji. Tous droits réservés.