Explore Projects
Find projects that match your skills and start building your execution portfolio
Find projects that match your skills and start building your execution portfolio
Find projects that match your skills and start building your execution portfolio
KBify is a fast and lightweight client-side utility platform that offers multiple tools (like PDF compression, image optimization, etc.) in one place β without any backend dependency. It is built using Vanilla JavaScript, uses reactive state management (ES6 Proxy), and supports PWA for offline usage.
DevPlace is a full-stack developer marketplace platform where developers can showcase and sell their digital products including websites, mobile apps, SaaS tools, and AI automation solutions. The platform provides a complete ecosystem for developers to monetize their work, while buyers can easily discover, purchase, and deploy ready-made solutions. It includes secure payments, product approval systems, analytics dashboards, and user-friendly interfaces for both buyers and sellers. DevPlace solves the problem of unused developer projects by turning them into revenue-generating assets and creating a centralized hub for high-quality digital products.
RockyAI is a smart AI automation system that can control both your browser and computer. It uses a mix of local logic, screen understanding, and cloud AI to perform tasks like clicking, typing, searching, and managing files automatically. It works fast for simple tasks and uses advanced AI (like Hugging Face) for complex ones. RockyAI can also learn from past actions, understand screen elements using screenshots, and be controlled through a real-time web dashboard.
GossipGo is a modern anonymous real-time chat platform where users can connect and communicate without revealing their identity. Built using modern web technologies like MongoDB, Express.js, Next.js, Node.js, and Socket.io, it features live messaging, smart matchmaking, chat history, and a powerful admin moderation system.
Real time problem faced by users to check that ### Problem Statement NFFEW=KEPWN F ### Solution Overview NFNPFFPNFPNMNFPN
An intelligent AI-powered solution that simplifies everyday tasks through automation and personalized assistance. ### Problem Statement People often rely on multiple tools and platforms to complete everyday tasks, find information, and solve problems. This creates unnecessary complexity, wastes time, and makes it difficult to access relevant information efficiently. There is a need for a simple, intelligent, and user-friendly solution that can understand user requirements and provide useful results quickly. ### Solution Overview Our project provides an AI-powered platform designed to simplify tasks through an intuitive interface and intelligent automation. Users can interact with the system using natural language, while the application processes their requests and generates relevant, contextual responses. The platform is built with a modern web architecture using React/Next.js for the frontend, Node.js for backend functionality, and AI APIs for intelligent processing. TailwindCSS is used to create a responsive and accessible user interface. Key features include AI-powered interactions, automated task handling, responsive design, fast processing, and an easy-to-use experience. The solution is designed to be scalable and can be extended with additional AI capabilities and third-party integrations.
Predict the true value of any gemstone using advanced machine learning. Fast, accurate and AI-powered valuation within milliseconds. This is a Production Level Machine learning Project
Knowledge Tracker β AI-Powered Smart Revision & Knowledge-Decay Prediction System Knowledge Tracker is an intelligent study companion that helps students revise more effectively by predicting when they are likely to forget a topic. Instead of relying on fixed timetables or random revision schedules, the system uses the Ebbinghaus Forgetting Curve to estimate each student's memory retention based on quiz performance, confidence level, topic difficulty, revision history, and the time elapsed since the last study session. The platform automatically prioritizes topics into High, Medium, or Low forget-risk categories and recommends exactly what the student should revise each day. Students can manually add topics or upload PDF notes, where an in-browser information retrieval pipeline extracts key concepts using tokenization, stop-word removal, heading detection, TF-based key phrase scoring, duplicate filtering, and OCR support for scanned documents. To reinforce learning, Knowledge Tracker generates AI-powered quizzes from tracked topics and updates the memory model dynamically based on student performance. Incorrect answers immediately increase the revision priority, enabling truly adaptive learning. A semantic search engine powered by sentence embeddings and a FAISS vector index allows students to search their uploaded study material by meaning rather than exact keywords. The platform also includes an AI study assistant that answers questions such as "What should I revise today?", explains memory retention for individual topics, recommends study resources, and summarizes learning progress. An interactive dashboard visualizes overall retention, study streaks, forget-risk distribution, and upcoming revision schedules, helping students stay consistent and maximize long-term knowledge retention. By combining cognitive science, artificial intelligence, semantic search, and personalized spaced repetition, Knowledge Tracker transforms traditional revision into a data-driven, adaptive learning experience that helps students remember more while studying less.
AutoResume AI is a modern, AI-inspired resume builder designed to help users create clean, professional, and ATS-friendly resumes in minutes. The platform focuses on simplicity, speed, and usability, allowing users to build resumes with live preview, smart suggestions, and instant PDF export. This project was built using AI-assisted development tools, demonstrating how ideas can be transformed into real-world products without traditional coding experience. It emphasizes strong UI/UX design, seamless authentication, and practical features required for job-ready resumes. The goal is to simplify resume creation while ensuring compatibility with Applicant Tracking Systems (ATS), increasing users' chances of getting shortlisted.
Aurastream is a modern full-stack video streaming platform that allows users to stream videos instantly using direct video links without downloading the complete file. The platform uses a proxy-based streaming architecture to bypass browser CORS restrictions and provide smooth playback with intelligent buffering. Users can paste a direct video URL and immediately watch the content through a modern responsive video player. Aurastream is designed with a clean dark-themed UI and optimized streaming flow where videos are loaded progressively in chunks instead of downloading the entire file at once. The system supports smooth playback, buffering, seeking, and responsive streaming across both desktop and mobile devices. The project demonstrates real-world concepts such as frontend-backend communication, proxy servers, streaming pipelines, deployment architecture, environment configuration, and cloud hosting. Aurastream is currently under development, and more advanced streaming, optimization, and AI-powered features are planned for future updates.
ShareJet is a modern cross-device sharing platform that allows users to instantly transfer text and files between devices using a temporary PIN system. Users can paste text or upload files, generate a unique PIN, and access the same content on another device by entering that PIN. The platform supports temporary cloud-based sharing with expiration timers and is designed for speed, simplicity, and accessibility. The project also includes an experimental βOptical QR Transfer Modeβ that uses dynamically animated QR codes and camera scanning for offline peer-to-peer communication between devices without internet, WiFi, Bluetooth, or NFC.
IoTMesh is an advanced ESP32-powered smart home automation ecosystem with real-time monitoring, Telegram integration, cloud alerts, voice command support, Firebase synchronization, weather prediction logic, and remote automation controls. The system allows users to monitor sensors, control appliances remotely, receive instant alerts, automate actions, and integrate Google Assistant voice commands using Telegram + IFTTT + cloud backend services. IoTMesh combines IoT hardware, cloud infrastructure, automation logic, and real-time communication into one scalable smart-home platform.
π Built an Interactive Gesture-Controlled 3D Cube using Python, OpenCV & MediaPipe! Excited to share one of my recent Computer Vision projects where I created a real-time hand gesture interaction system that allows users to create, move, and interact with a virtual 3D cube using only hand movements. πΉ Features: β Real-time hand tracking using MediaPipe β Dynamic hand landmark visualization β Gesture-based cube creation using pinch detection β Static cube persistence after gesture completion β Drag-and-move interaction with hand gestures β βSlap to Rotateβ mechanism using motion velocity detection β Dynamic visual effects with adaptive colors
Campus Marketplace is a full-stack web app for buying and selling items within a college community. It solves the chaos of WhatsApp-based selling by providing a centralized platform with product listings, real-time chat, and secure user interaction.
Morphr is an AI-powered resume optimization platform that automatically tailors resumes to specific job descriptions. It solves a common problem faced by students and job seekers β the need to manually rewrite resumes for every application to improve ATS (Applicant Tracking System) compatibility.
Sage is an advanced AI-powered research assistant designed to improve the reliability of Retrieval-Augmented Generation (RAG) systems. Unlike traditional βchat with PDFβ tools, Sage implements a multi-stage retrieval pipeline to significantly reduce hallucinations and improve answer accuracy. The system combines hybrid search (semantic vector search + keyword-based retrieval) with an LLM-based re-ranking layer powered by Gemini 2.5 Flash. This ensures that only the most relevant and high-quality context is used for generating responses. Sage also introduces agentic capabilities, allowing the system to autonomously decide when to fetch external data. It can perform real-time web searches using Serper, scrape and extract content from web pages using Firecrawl, and summarize long-form documents intelligently. The project demonstrates the evolution from simple prompt-response systems to intelligent AI agents capable of tool usage, contextual reasoning, and dynamic decision-making.
HeartGuard AI is a medical chatbot that helps patients and healthcare providers with heart health questions. It uses a team of AI agents backed by real medical databases to provide personalized cardiac health guidance.
PROBLEM: Smart Screen Time Monitoring & Reduction System Problem Statement Nowadays people spend excessive time on mobile phones, social media, and games. This leads to several issues such as: β Reduced productivity β Sleep disturbances β Eyestrain and health issues β Mental stress and reduced focus An intelligent system is required to monitor screen usage and encourage healthier digital habits. SOLUTION: DigiWell is a full-stack digital wellness platform for tracking screen time, analyzing usage behavior, and improving daily focus habits. This repository includes: A Flask backend API with analytics, planner, mood, and AI endpoints A React + Vite frontend dashboard (digiwell/) Background app-usage trackers for desktop/browser activity SQLite storage with migrations ML training/inference utilities for risk and productivity predictions
PROBLEM: Create an assistant that helps users understand the election process, timelines, and steps in an interactive and easy-to-follow way. SOLUTION: VoterAssist is an AI-powered educational web app designed to guide "First-Time Voters" through the 2024 Indian General Elections. Key Features: Interactive Timeline: A step-by-step guide from checking eligibility to voting day. Gemini AI Chat: A smart assistant trained on ECI guidelines to answer voter questions. Data Dashboard: Visualizations of historical election results across states. Polling Booth Finder: A mock database search with Google Maps integration. Tech Stack: React (Vite), Tailwind CSS, Python Flask, and deep integration with 6 Google Services (Gemini, Maps, Analytics, Translation, Cloud Run, Calendar). It was built to achieve 100% in security, accessibility, and testing metrics.
This project presents a complete AI-based loan default prediction system for banking applications. It uses both machine learning and deep learning models to identify borrowers who are likely to default on loans. The system includes data cleaning, preprocessing, class imbalance handling using SMOTE and SMOTE-Tomek, model training, and explainable AI techniques like SHAP and LIME for transparent decision-making. Advanced ensemble models based on DenseNet and ResNet achieved up to 99.2% precision, improving prediction accuracy, recall, and reliability for real-world banking risk assessment.