Projects

A command-line tool that turns any YouTube video or playlist into structured, hierarchical Markdown study notes automatically. The Problem YouTube is one of the richest learning platforms available — lectures, conference talks, tutorials, full courses. But video is a passive medium. You watch, and two days later the details are gone. Writing notes by hand doubles your time investment. NoteWise eliminates that tradeoff entirely. What It Does Point NoteWise at any YouTube URL and it produces deep study notes with headers, sub-topics, definitions, and examples — not a transcript dump. Chapter-aware videos get per-chapter note files. Long playlists are processed concurrently and organized automatically. Single videos, full playlists, and batch URL files. Chapter detection with separate notes per chapter. Optional multiple-choice quiz generation. Raw transcript export in plain .txt or timestamped .json. Local SQLite cache so already-processed videos are skipped. 8+ LLM providers via LiteLLM — Gemini, OpenAI, Anthropic, Groq, and more. Docker-ready for reproducible runs. Why I Built It I needed a way to catch up on university lecture recordings and conference talks without watching everything twice. NoteWise started as a personal script and grew into a fully packaged, CI-tested open-source tool.

An AI-powered resume content generator that turns your GitHub repositories into professional, ATS-optimized resume sections — tailored to the exact job you're applying for. The Problem Most developers undersell their work on resumes. The gap isn't capability — it's knowing how to translate a codebase into achievement-driven language that passes ATS filters and impresses hiring managers. GitResume closes that gap automatically. What It Does Paste a repository URL and a job description. GitResume analyzes the codebase using Tree-sitter, extracts meaningful technical signals, and generates resume content aligned to the role — not generic bullet points, but keyword-matched, achievement-focused writing grounded in your actual code. Smart code analysis that detects tech stacks, project structure, and quantifiable metrics. Job-tailored bullet points optimized for ATS with relevant keywords. Interview question generation with comprehensive answers based on your project. Real-time streaming UI via WebSockets showing live progress through the pipeline. GitHub OAuth for secure access to private repositories. Modular AI backend supporting Gemini, OpenAI, Groq, and Claude. What It Generates Professional project titles and descriptions, achievement-focused bullet points with quantifiable results, full tech stack breakdowns, technical setup documentation, improvement suggestions, and interview preparation materials.

A secure, full-stack personal finance application that gives you complete visibility and control over your income, expenses, and budgets — with real-time charts and detailed reporting built in. What It Does Finance Manager brings all your financial activity into one clean interface. Track transactions, set category budgets, monitor spending patterns, and project cash flow — without trusting a third-party service with your financial data. An interactive dashboard gives you a real-time snapshot of income, expenses, and balance with visual income-vs-expense comparisons. Transaction management supports payment methods, tags, and notes for full context on every entry. Smart budgeting lets you set monthly limits per category with progress indicators and spending alerts. Dynamic reports surface monthly trends, category breakdowns, and cash flow projections. Light and dark themes, multi-currency support, and custom date formats round out the experience. Security Built with security as a first-class concern. Passwords are hashed with Bcrypt. Authentication endpoints are rate-limited against brute-force attacks. Sessions use HttpOnly cookies to prevent XSS. Sensitive admin actions require password confirmation. Architecture Clean separation between a RESTful Flask API and the frontend interface. MongoDB's flexible schema handles efficient transaction storage, and the entire stack runs in Docker Compose for straightforward deployment.

A real-time mission control desktop application built for the CanSat India competition. The system bridges hardware telemetry with an intuitive GUI — providing complete satellite oversight from launch to recovery. The Context CanSat is a competition where teams design, build, and launch a satellite the size of a soda can. The ground control station is the nerve center of the operation — everything the team sees, commands, and records during the mission flows through it. What It Does A live telemetry dashboard renders real-time altitude, pressure, and sensor readings through dynamic charts and custom gauge widgets. An integrated GPS mapping widget tracks the satellite's position as it updates. A mission control panel handles one-click command transmission for critical events like payload deployment, with instant feedback. All telemetry is automatically logged to timestamped CSV files for post-flight analysis using Pandas and Matplotlib. Under the Hood A dedicated serial handler runs on a background thread, continuously reading telemetry packets from ground station hardware without blocking the UI. PyQt5's signals and slots pattern keeps GUI updates thread-safe and the interface responsive under load — critical when mission data is streaming in real time. Impact Delivered a complete IoT application covering live monitoring, remote command execution, and post-flight analysis tools. The system's stability and reliability were directly tied to mission success.

A complete ground control and telemetry solution built for the Teknofest 2024 Model Satellite Competition. The system spans two platforms — a feature-rich desktop ground station and a Raspberry Pi satellite payload — connected over WebSockets for real-time flight data collection, transmission, and visualization. The Context Teknofest is one of the largest aerospace and technology competitions in the world. The model satellite category requires teams to design both the satellite payload and the ground control system that monitors it. Every sensor reading, orientation shift, and GPS coordinate during flight flows through this system. What It Does A real-time telemetry dashboard renders live altitude, pressure, temperature, and GPS data using PyQtGraph. A Folium-based interactive map plots the satellite's trajectory with live polyline updates as coordinates stream in. A VTK-powered 3D widget visualizes roll, pitch, and yaw in real time — giving the team immediate spatial awareness of the satellite's orientation mid-flight. All telemetry is automatically logged to CSV for post-flight blackbox analysis. A standalone CSV simulator allows full ground station testing without a live satellite connection. The Satellite Payload The Raspberry Pi payload interfaces directly with environmental and orientation sensors — BMP390 for altitude and pressure, TMP117 for temperature, and BNO055 for 3D orientation — via Adafruit CircuitPython libraries. A Flask server on the Pi handles log access and media streaming alongside the telemetry pipeline. Architecture Dual-platform design with async WebSocket communication handling concurrent sensor streams and ground station updates. The data pipeline flows continuously from sensor acquisition through transmission, visualization, and logging — with no blocking operations on either end.