Zahid Hasan TonmoyZahid Hasan TonmoyZahid Hasan Tonmoy

lexPath is a mobile-first gig economy platform designed specifically for Bangladesh. It helps connect job seekers (like students, freelancers) with employers (such as small businesses or households) who need temporary, flexible workers for short-term tasks. Developed using Flutter (frontend) and Supabase (backend), FlexPath provides a localized, secure, and user-friendly job matching experience. The app ensures trust and authenticity through NID (National ID) verification, real-time messaging, and role-based dashboards for different user types. ⚙️ Core Functionalities: 👤 User Roles & Access Job Seeker Panel: Register, upload NID, browse/apply to jobs, chat with employers. Employer Panel: Post jobs, review applicants, hire workers, message seekers. Admin Panel: Approve NID submissions, manage users, resolve disputes. 🔐 Authentication & Verification Secure email/password & social logins NID verification with admin approval Role-based access control 📲 Job System Dynamic job feed with search & filters Job posting with details like title, description, budget Job application & status tracking Ratings & reviews after job completion 💬 Real-Time Chat In-app messaging system using Supabase Realtime Instant communication between seekers and employers 🛡️ Admin Features Dashboard with user/job stats NID approval panel User and job management Manual dispute resolution tools 💼 Technology Stack Frontend: Flutter 3.x Backend: Supabase (PostgreSQL, Auth, Storage, Realtime) Dev Tools: VS Code, Android Studio, GitHub, Figma

This project focuses on predicting the daily closing price of gold using various machine learning regression algorithms. Based on historical financial data like Open, High, Low, and Volume, the study aims to determine the most accurate model for gold price forecasting. The project also includes a web application that allows users to input data and receive real-time price predictions using the trained models. The goal is to provide a practical tool for investors, students, and researchers to make informed predictions using machine learning. ⚙️ Core Functionalities: 📥 Data Preprocessing Load and clean historical gold price dataset Convert date formats, handle missing values (none found) Apply feature scaling (StandardScaler) 📊 Exploratory Data Analysis (EDA) Correlation heatmaps and pairplots Time series visualizations of price trends Feature relationship analysis (Price vs. Open, High, Low, Volume) 🔧 Model Training & Evaluation 15+ regression models tested: Linear Regression, Ridge, Lasso, Random Forest, Gradient Boosting, Huber Regressor, etc. Metrics used: R², MSE, RMSE, MAE Best model: Huber Regressor with R² ≈ 0.9999 🤖 Prediction System Predict prices using new input values (Open, High, Low, Volume) Save results to CSV 🌐 Web Application Built to let users interactively input data Select preferred algorithm from dropdown View predicted gold price instantly 📈 Performance Insights Extremely high accuracy using price-related features Demonstrated suitability for "nowcasting" daily prices

Halarnati is a lightweight web-based file and text sharing platform designed with a playful, user-friendly interface. Created by Zahid Hasan, it allows users to upload files, write descriptions, and share content securely or publicly. The interface is styled with modern UI elements and includes casual Bangla-English mixed language to make it more relatable, especially for Bangladeshi users. Functionalities: Upload files with a title and text description Optionally add a password to lock entries Search for uploaded entries View and download uploaded files Copy shared links and content to clipboard Paginate through multiple entries Password-protected content unlocking feature

OffenseOrbit is an advanced crime management platform designed to make communication between citizens and law enforcement seamless. By harnessing technology, OffenseOrbit helps streamline processes and ensure public safety.

Smart Drainage Monitor Project Description The Smart Drainage Monitor is an innovative IoT-based system designed to prevent urban flooding by monitoring drainage chambers, detecting blockages, and automating clearance. This project addresses the critical issue of clogged drains, which often lead to waterlogging in cities, causing damage and disruption. By combining advanced hardware, cloud technology, and a user-friendly Android app, the system ensures real-time monitoring and efficient management of drainage systems. Project Overview The Smart Drainage Monitor uses an ESP32-S3 microcontroller to collect data from multiple sensors, including water level, sonar (HC-SR04), gas (MQ8), temperature (LM35), IR, flame, and GPS modules. These sensors act as the system's "eyes," detecting water levels, potential blockages, and environmental conditions. The data is sent to Firebase Realtime Database, a cloud-based "notebook" that stores and updates sensor readings instantly. The Android app, developed as the primary user interface, displays this data, alerts users to blockages, and allows manual or automatic control of a servo motor to clear drains. Key Features Real-Time Monitoring: The app shows water levels for three drainage chambers, updated every 2 seconds. Blockage Detection: If two chambers are full and one is empty, the empty chamber is flagged as blocked. Sensor Insights: Displays data from sonar, gas, temperature, and GPS. User Controls: Includes a map button, refresh button, and servo switches for manual or auto mode. Technical Components Hardware: ESP32-S3 with sensors running MicroPython to collect and send data to Firebase. Software: Android app (Java, minSdk 26) with Firebase integration. Cloud: Firebase Realtime Database stores sensor data and servo control settings.

Curious Cart BD is a full-stack e-commerce platform built with Next.js, React, and TypeScript. This isn't just another online store — it's a thoughtful solution that combines technology with emotional wellness. Key Features: - Responsive UI/UX with Tailwind CSS - Full cart & wishlist functionality with React Context API - Product management system with dynamic variants - Social sharing capabilities - Interactive image slideshow - RESTful API endpoints for product data - Secure authentication system Tech Stack: Next.js, React, TypeScript, Tailwind CSS, REST API

Jerseyvault is a full-featured e-commerce platform built with React.js and powered by Supabase. It showcases full-stack development skills and attention to detail in creating a seamless user experience. Key Features: - Frontend: React.js with React Router for SPA navigation - Backend: Supabase (PostgreSQL, Authentication, Real-time APIs) - State Management: React Context API for cart functionality - Responsive UI: Custom CSS with mobile-first design - Complete shopping flow: Browse → Add to Cart → Checkout - User authentication and profile management - Order tracking and history - Wishlist functionality - Product reviews and ratings system - Inventory management with stock validation

Vortex Shield is a cybersecurity-themed file encryption suite featuring a distinct cyberpunk aesthetic. It provides tools for securely encrypting, decrypting, and permanently deleting files. The project consists of two main components: a native desktop application for Windows and a modern web application. Desktop Application (Python): - GUI Framework: CustomTkinter with Drag-and-Drop support - Cryptography: AES-GCM algorithm, Argon2id key derivation - Features: File encryption (.vortex extension), decryption, and secure deletion (3-pass overwrite) - Duress mode, sound effects, single-file executable via PyInstaller Web Application (React/Vite): - Tech Stack: React, TypeScript, Vite - UI Components: CyberButton, MatrixText, LogViewer, DropZone - Client-side encryption simulation in a browser

LocalDrop Pro is a modern, secure, peer-to-peer (P2P) file sharing and real-time chat web application. It allows devices to connect directly via WebRTC to transfer files and messages without intermediate servers. Key Features: - Direct P2P Transfer via WebRTC (PeerJS) - Easy Pairing via 6-digit short ID or QR code - Universal file support: any type and size - Directory support: auto-zips folders before sending - Drag & Drop interface - Queue system with pause, resume, and cancel - Real-time progress bars and transfer history - Integrated text chat between peers - Multiple visual themes: Aurora, Cyberpunk, Minimal - PWA support for native-like installation - End-to-End Encryption via AES-GCM