// development blueprints
My Projects
Hyphenet Enterprise Portal
Node.js, Express & MySQL Relational Architecture
A unified, full-stack enterprise web portal and IT service desk management system developed for Hyphenet Enterprises. Delivers a 5-role support desk with real-time SSE event streaming, automated SLA escalations, and Google Workspace email routing, integrated with a comprehensive Staff HRMS & Automated Payroll Engine and Multi-Hub Warehouse Inventory Tracking.
Hyphenet Enterprise Portal
A unified enterprise IT rentals platform and 5-role support helpdesk. Features real-time SSE ticket queues, Google Workspace email routing, Staff HRMS & Automated Payroll Engine, and Multi-Hub Warehouse Inventory Tracking.
Hyphenet Enterprise Portal | Hyphenet Enterprise Portal
A unified, full-stack enterprise web portal and IT service desk management system developed for Hyphenet Enterprises. Unites an 18-page commercial catalog and 9-page statutory legal hub with a multi-tenant IT service desk (/portal), real-time Server-Sent Events, automated SLA escalations, Google Workspace email routing, Staff HRMS & Automated Payroll Engine, and Multi-Hub Warehouse Inventory Tracking.
Key Features:
- Multi-Role IT Service Desk (/portal) with dedicated operational consoles for Enterprise Admin, Service Coordinator, Field Engineer, Client IT Support, and Individual Clients
- Real-Time Server-Sent Events (SSE) on /api/realtime/stream for zero-latency dashboard queue synchronization with automated 20-second keep-alive heartbeats
- Automated SLA Tracking & Escalation Engine with dynamic timers, 24-hour hardware replacement shifts, and background Team Lead email alerts
- Staff HRMS & Automated Payroll Engine with employee attendance tracking, leave approval workflows, and automated salary slip generation
- Multi-Hub Warehouse Inventory Tracking managing enterprise servers, laptops, and networking stock with hardware fault replacement workflows
- Google Workspace Service Account Integration (Gmail REST API v1) with domain-wide delegation for direct transactional email dispatch from support@hyphenet.net
- ExcelJS Bulk Incident Stream processing hundreds of rows per spreadsheet with external ticket duplicate prevention
- Progressive Web App (PWA) with Service Worker offline caching (portal/sw.js) and statutory DPDP Act 2023 cookie consent management
System Architecture:
High-performance Node.js & Express 4.x RESTful backend paired with a self-healing MySQL / MariaDB database (automated runtime schema integrity, composite indexes, and utf8mb4 collation). Pure Web Standards frontend (Semantic HTML5, CSS3 design tokens, modular Vanilla JavaScript, PWA) deployed on Hostinger Cloud with LiteSpeed/Apache .htaccess rules, PM2, and a 50-category defensive security posture.
Technologies & Tools: Node.js, Express.js, MySQL, Server-Sent Events, PWA, Google Workspace API, Hostinger Cloud, ExcelJS, JWT
Fluxify Media Platform
Production-ready PHP 8.2 & MySQL 8 web platform and custom admin CMS for digital agencies with dynamic case studies, blog management, and PHPMailer SMTP integration.
Fluxify Media Platform | Fluxify Media Platform
A production-ready full-stack website and custom administrative CMS developed for Fluxify Media. Built with modern PHP 8.2 and MySQL 8, it powers agency service showcases, client case studies, testimonials, rich blog publishing, and secure lead generation using PHPMailer.
Key Features:
- Custom Admin CMS Control Panel (/admin/*) with secure session management
- Dynamic Case Studies, Portfolio & Testimonials publishing system
- PHPMailer Gmail SMTP server delivery engine with dedicated inquiry routing
- Rich Content authoring suite with Quill Editor and DataTables integration
System Architecture:
PHP 8.2 backend with MySQLi prepared statements and MySQL 8 database. Frontend with Tailwind CSS, AOS scroll animations, and Quill rich-text editor.
Technologies & Tools: PHP 8.2, MySQL 8, PHPMailer, Tailwind CSS, JavaScript ES6, Quill Editor, DataTables
CampusPulse LMS
Full-Stack Learning Management System for B.N. College Computer Applications Dept. Features OTP verification, Cropper.js avatar checks, MCQ testing, and PWA Web Push notifications.
CampusPulse LMS | CampusPulse LMS
A full-featured academic Learning Management System engineered specifically for the Department of Computer Applications at B.N. College. Streamlines institutional workflows with student self-registration, teacher approvals, courseware distribution, and online examinations.
Key Features:
- Student self-registration with OTP verification via Brevo HTTP Email API
- Mandatory profile picture square-cropping via Cropper.js with teacher approval queue
- Semester-wise courseware catalog and lecture note distribution
- Timed online MCQ examination engine with automatic evaluation and ranking
- Daily attendance tracking with automated 75% statutory threshold alert warnings
- Progressive Web App (PWA) with VAPID Web Push Notifications via Service Workers
System Architecture:
Node.js & Express.js REST API with MongoDB Atlas and Mongoose ODM. HTTPOnly JWT cookies, bcryptjs password hashing, and express-rate-limit protection.
Technologies & Tools: Node.js, Express.js, MongoDB Atlas, Brevo API, Cropper.js, PWA, JavaScript ES6
Vistredge Site
Designed and built the corporate website presence for Vistredge, ensuring high accessibility performance standards and seamless responsive animations.
Vistredge Corporate Site | Vistredge Corporate Site
A modern corporate web presence and digital services showcase built with clean semantic standards. Highlights brand storytelling, IT consulting capabilities, client feedback, and lead conversion.
Key Features:
- Clean multi-page architecture (Home, Services, Testimonials, Contact)
- High-accessibility standards, optimized loading speeds, and smooth mobile UX
- Lightweight responsive interface adhering to cyber-minimalist principles
System Architecture:
Semantic HTML5 and modular responsive CSS3 with vanilla JavaScript DOM interactions.
Technologies & Tools: HTML5, CSS3, JavaScript ES6, PHP, Responsive UI
StrengthFuel
An AI-powered sports nutrition and recovery web application for competitive weightlifters. Features natural language Indian food parsing, MPS meal analysis, and a 4-layer anti-doping supplement inspector.
StrengthFuel | StrengthFuel
An intelligent AI-powered sports nutrition and recovery web application designed for weightlifters and power athletes. Combines sports science principles with Google Gemini AI to provide granular macronutrient tracking, meal quality analysis, and substance safety screening.
Key Features:
- Natural Language Indian Food Parser supporting local staples (roti, dal, paneer, eggs, rice) with macro breakdown
- Analyse My Meal tool evaluating Muscle Protein Synthesis (MPS) thresholds
- 4-Layer Anti-Doping Supplement Inspector checking substances against banned lists (WADA/NADA standards)
- Body weight and recovery trend tracking with dynamic SVG moving average charts
- Personalized Coach AI Assistant powered by Google Gemini API grounded in live athlete metrics and daily briefings
System Architecture:
Multi-page HTML5/CSS3/Vanilla JS architecture with hybrid client router, powered by Firebase Cloud Functions (Node.js) invoking Google Gemini API with sports-science fallbacks, and Firestore DB caching.
Technologies & Tools: JavaScript ES6, Firebase Functions, Google Gemini AI, Firestore, HTML5, CSS3, PWA
mdtopdf | Markdown Studio
100% client-side Markdown to vector PDF & DOCX converter with glassmorphic dark/light UI, side-by-side live preview, auto-save, syntax highlighting, and custom presets.
mdtopdf | Markdown Studio | mdtopdf | Markdown Studio
A modern, high-speed, and completely client-side Markdown editor and document converter featuring a sleek Obsidian & Yellow glassmorphic interface. Enables real-time side-by-side editing, live typography presets, and instant vector PDF or MS Word (.docx) export with zero server uploads.
Key Features:
- Dual-pane split workspace with synchronized scrolling and line number gutters
- 4 Document styling presets: Dark Yellow (Obsidian), GitHub Style, Academic Paper (Serif), and Minimalist
- Floating formatting toolbar (headings, code blocks, tables, task lists, undo/redo)
- Live document metrics: Real-time word count, character count, line numbers, and estimated reading time
- Multi-format Export dialog: Crisp vector PDF compilation with high-contrast text and native Word OpenXML (.docx) export
- LocalStorage auto-saving with live status indicator and drag-and-drop file upload
System Architecture:
100% client-side compilation using Marked.js for markdown parsing, DOMPurify for XSS sanitization, Highlight.js for syntax highlighting, html2pdf.js for vector PDF generation, and html-docx-js for OpenXML Word conversion.
Technologies & Tools: HTML5, CSS3 Glassmorphic, JavaScript ES6, Marked.js, DOMPurify, html2pdf.js, html-docx-js
ToolHub Pro
All-in-One client-side web utilities platform packing 120+ tools for images, PDFs, and converters. Full PWA offline support, smart global search, and jet-black dark mode.
ToolHub Pro | ToolHub Pro
A comprehensive and completely private web tools platform offering 120+ utilities across image manipulation, PDF management, and file format conversions. Operates 100% client-side in browser memory, ensuring user files and data never leave the local device.
Key Features:
- 100% Client-Side Sandbox: Zero server uploads, zero telemetry, immediate browser memory cleanup
- PWA Full Offline Support: Service worker pre-caches all 120+ tools for complete offline execution
- Smart Global Search: Keyboard-accessible floating search bar (ArrowUp/ArrowDown/Enter/Escape)
- Image Utility Suite: Crop, resize, compress, aspect ratios, OCR text recognition, filters, and meme creation
- PDF Utility Suite: Merge, split, annotate, protect, sign, encrypt, and convert document formats
- Encoders & Converters: Base64, Hex, Binary, URL, Hash, JSON/XML, and Unit/Currency converters
System Architecture:
Modular ES6 JavaScript architecture using native Browser APIs (Canvas API, Web Workers, File API) packaged with a Progressive Web App (PWA) service worker caching layer.
Technologies & Tools: JavaScript ES6, PWA, Service Worker, LocalStorage, HTML5, CSS3
DB File Explorer
A modern, fast, and completely client-side database exploration and SQL management tool running in the browser on SQLite WebAssembly (SQL.js). Inspect schemas, run queries, filter and edit records, and export datasets with zero server uploads.
DB File Explorer | In-Browser SQLite WebAssembly Manager
A modern, fast, and completely client-side database exploration and management tool running directly in your web browser. Built on SQLite WebAssembly (SQL.js), DB File Explorer empowers developers, data analysts, database administrators, and students to inspect schemas, execute SQL queries, filter and edit records, and export datasets without installing native desktop software, running database engines, or setting up backend servers. 100% private and offline-first: your databases and SQL scripts are processed exclusively within local browser memory with zero external server transmissions.
Key Features:
- In-Browser SQLite WebAssembly Engine: Full SQLite 3 SQL query execution powered by SQL.js running entirely in local browser memory
- Comprehensive File Support: Ingests native SQLite databases (.db, .sqlite, .sqlite3, .db3) with magic header checks and runs .sql dumps with automated MySQL/MariaDB dialect cleanup
- Database Schema Inspector: Visual breakdown of tables, columns, data types, primary keys, foreign key constraints, indexes, views, and live row counts
- Interactive Tabular Data Grid: Paginated record browsing with custom page sizes, column sorting, column visibility toggles, and BLOB/NULL indicators
- Multi-Column Filter System: Multi-condition filtering supporting text operators (Contains, Starts with, Exact) and numeric operators (=, >, <, Between, NULL)
- Interactive SQL Console: Monospace query editor with instant execution (Ctrl+Enter), microsecond runtime telemetry, predefined snippets, and session query history
- Record CRUD Editor: Safe record insertion and row editing modals dynamically generated from PRAGMA table_info, with rollback to original state and unsaved changes tracking
- Comprehensive Export Suite: Multi-format exports including RFC 4180 CSV, formatted JSON, multi-sheet Excel (.xlsx) with formula injection defense via SheetJS, and binary .db file downloads
System Architecture:
Zero-build static client-side web application built with Vanilla HTML5, CSS3, and ES6 JavaScript. Ingests databases via FileReader and Uint8Array into an in-memory SQLite 3 WebAssembly engine (SQL.js), rendering reactive tabular data grids with SheetJS spreadsheet generation and zero server-side infrastructure.
Technologies & Tools: SQLite WebAssembly, SQL.js, JavaScript ES6, SheetJS (xlsx), HTML5, CSS3, Web Workers
Expenses Tracker App
Cross-platform mobile application designed for budgeting, monitoring income flow, and displaying structured graphical summaries of weekly/monthly expenses.
Expenses Tracker App | Expenses Tracker App
A cross-platform mobile application developed with React Native for personalized finance management. Empowers users to record, categorize, monitor, and search their income streams and expense transactions with custom multi-source options and structured graphical summaries.
Key Features:
- Personalized income logging with configurable multi-source options (Salary, Freelance, Investments, etc.)
- Categorized expense tracking with custom tags, payment modes, and timestamped notes
- Fast search, category filters, and historical financial transaction queries
- Structured weekly and monthly financial summaries with visual breakdown reports
- Local offline database storage ensuring complete user privacy and instant responsiveness
System Architecture:
Cross-platform mobile client built using React Native and JavaScript ES6, utilizing local database storage for offline-first data persistence.
Technologies & Tools: React Native, JavaScript ES6, Android Studio, Local DB, Mobile UI
Plant Diseases Detector
A deep learning model utilizing Convolutional Neural Networks (CNN) trained to identify plant leaf diseases from user-uploaded images, helping farmers diagnose crops.
Plant Diseases Detector | Plant Diseases Detector
An AI-driven agricultural diagnosis system that employs Deep Learning Convolutional Neural Networks (CNN) to detect and classify plant leaf diseases from user photos. Trained in Python/TensorFlow and converted to TensorFlow.js for instantaneous client-side browser inference without requiring backend GPUs.
Key Features:
- Deep Learning CNN architecture trained on diverse crop leaf disease datasets
- Instant in-browser disease diagnosis using TensorFlow.js client-side inference
- Image upload and live photo analysis with real-time class probability scores
- Detailed diagnosis breakdowns with actionable treatment and crop remedy recommendations
- Multi-crop coverage (Tomatoes, Potatoes, Corn, Apples, and more)
System Architecture:
Trained with Python, TensorFlow, and Keras on Google Colab, converted to TensorFlow.js graph models (tensorflowjs-model/) for client-side JavaScript execution in the browser with zero server latency.
Technologies & Tools: Python, TensorFlow, TensorFlow.js, Keras, CNN, JavaScript, HTML5/CSS3
Fake News Detector
An NLP semantic classification system analyzing news text patterns. Employs Sentence Transformers (all-MiniLM-L6-v2) and trained classifiers with Streamlit interface.
Fake News Detector | Fake News Detector
An advanced Natural Language Processing (NLP) system designed to detect and classify deceptive or fabricated news text. Utilizes pre-trained Sentence-Transformers (all-MiniLM-L6-v2) to extract dense semantic embeddings from news headlines and body text, classified via trained ML models on the 72,000+ article WELFake dataset with an interactive Streamlit UI.
Key Features:
- Dense Semantic Sentence Transformers embedding extraction using all-MiniLM-L6-v2
- High-accuracy classifier model trained on the comprehensive WELFake dataset (72,000+ articles)
- Real/Fake classification with real-time probability confidence score breakdown
- Feature scaling pipeline with persistent .pkl model artifact caching for fast predictions
- Interactive Streamlit web interface (app.py) for live text pasting and immediate verification
System Architecture:
Python NLP pipeline using Sentence-Transformers for semantic vectorization, Scikit-Learn classifier with feature scaling, and Streamlit for reactive web application rendering.
Technologies & Tools: Python, Sentence-Transformers, Scikit-Learn, Streamlit, Joblib, NumPy, NLP