Our project aims to tackle the challenges posed by deepfake technology to media authenticity by leveraging the combined power of blockchain and machine learning. Deepfake Detection: Uploaded media is analyzed by three backend nodes running a deep learning model to assess if it's likely a deepfake. A consensus mechanism can be used for higher reliability. Integrity Verification: A cryptographic hash (SHA-256) of the original media file is calculated before upload. Decentralized Storage: Verified authentic media is uploaded to IPFS via Pinata, ensuring content-addressable, decentralized storage. Immutable Record: The IPFS CID (Content Identifier) and the calculated hash are stored immutably on an Ethereum-compatible blockchain using a Solidity smart contract. Secure Sharing & Traceability: The smart contract manages fine-grained, item-level sharing permissions. Sharing actions are logged via events, enabling traceability of who shared what with whom. Client-Side Verification: When viewing media, the application fetches the content from IPFS, recalculates its hash, and verifies it against the hash stored on the blockchain, ensuring tamper-evidence.

Last year, I built a movie platform where anyone can download or stream movies without being interrupted by numerous ads. Users can access content by paying a small fee, saving both time and protecting their privacy — all without the need to create an account. The main goal of this platform is to make it easy to find and enjoy both new and classic movies securely, without the risk of personal data leaks. It’s built using React.js and cloud storage for efficient and scalable movie hosting.

I developed a full-stack Industrial IoT framework that integrates Python-based Digital Twins, simulated embedded firmware, and a native Android monitoring application. The system models industrial devices through digital twins while communicating with simulated microcontroller firmware built for the ESP32 using the Wokwi simulator. A native Android application built using Android Studio provides real-time telemetry visualization, device control, and monitoring. The architecture enables real-time synchronization between the simulated physical device and its digital twin. Sensor data generated by the firmware is transmitted to the Python backend, where the digital twin processes, analyzes, and stores the state of the system. The Android application retrieves telemetry data and displays it through an intuitive interface for monitoring and interaction. Industrial IoT development often faces three major challenges: 1. Lack of accessible testing environments Developing embedded IoT systems typically requires physical hardware, which increases cost and slows experimentation. 2. Difficulty in validating system behavior before deployment Without a digital representation of devices, it is difficult to simulate scenarios, monitor internal states, and test system responses safely. 3. Fragmented development pipelines Firmware, backend processing, and user interfaces are often developed separately, making integration complex. Key Technical Components Embedded Layer: Simulated ESP32 firmware generates sensor telemetry and device status data. Digital Twin Layer: Python services maintain virtual device models, process telemetry streams, and manage device state. Mobile Telemetry Layer: A native Android application visualizes live device data and provides control interfaces. Communication Layer: The system enables real-time communication between firmware, digital twin services, and the mobile application. Challenges Faced 1. Synchronizing Physical and Digital States Maintaining consistent state between simulated firmware and its digital twin required careful design of data pipelines and update mechanisms. 2. Real-Time Telemetry Handling Processing streaming sensor data while maintaining responsiveness in the mobile application required efficient backend handling and structured APIs. 3. Cross-Layer Integration Ensuring smooth communication between firmware simulation, backend services, and the mobile app required careful protocol design and modular architecture. 4. Simulating Realistic Device Behavior The firmware simulation needed to replicate realistic sensor outputs and device responses to properly test system logic.

### Overview I built a **full-stack online coding platform**, inspired by LeetCode, to help users practice data structures and algorithms efficiently. The platform allows users to solve problems in multiple languages, track their progress, and experiment in a custom coding playground. ### Problem Solved Many learners struggle to find a centralized platform to practice DSA with real-time code execution and progress tracking. This platform addresses that by combining problem-solving, tutorials, and performance analytics in one place. ### Key Features * **User Authentication & Authorization:** Secure login/registration with JWT and role-based access. * **Dynamic Problem Library:** Problems fetched from GitHub for easy updates and modularity. * **Multi-language Code Execution:** Solve problems in JavaScript, Python, C++, and Java using Monaco Editor integrated with Judge0 API. * **Progress Tracking:** Dashboard shows solved problems, success rates, and category-wise performance. * **Playground:** Test custom code snippets with user-defined inputs, stored in local storage. ### Challenges & Learnings * Integrating **Judge0 API** for secure, real-time multi-language code execution. * Designing a **scalable MERN architecture**. * Handling **dynamic problem fetching** from GitHub while maintaining performance. * Creating a user-friendly interface with **Monaco Editor** and persistent code storage. ### Outcome The platform now allows users to practice coding problems seamlessly, track performance, and learn algorithms systematically, bridging the gap between learning and application.

I built a Decentralized IoT Data Sharing Marketplace where IoT devices can securely share and monetize sensor data. Each dataset is tokenized as an NFT to ensure ownership, authenticity, and traceability. Smart contracts enforce role-based access control for providers and buyers. The project solves trust and ownership issues in traditional IoT data sharing and helped me gain hands-on experience with Solidity, Hardhat, Ethers.js, and React.js.

AI-Powered Resume Analyzer • Built a full-stack resume analysis system using Python, Streamlit, and MySQL to streamline and automate the hiring process. • Implemented NLP techniques using NLTK and SpaCy for resume text extraction, named entity recognition (NER), and structured parsing of skills, experience, and education. • Designed an interactive recruiter dashboard with Plotly visualizations to analyze candidate skill distributions, experience levels, and talent pools. • Developed a secure relational database using MySQL and PyMySQL to store resumes, extracted entities, and recruiter insights efficiently. • Integrated a recommendation engine to suggest skill improvements and career upskilling paths based on resume gaps and industry trends.