All Projects

A live-fire training ground for terminal mastery. Users solve real-world Linux puzzles in isolated Docker environments with namespace isolation, practicing everything from symlinks to process debugging. Built the infrastructure to give each user their own safe terminal sandbox – basically mini AWS for learning.

An AI agent that autonomously performs SDE-1 level tasks including GitHub operations (commit, PR, review), Slack/email handling, and voice interactions through natural language. Features a CLI tool for local task automation.

A hackathon-winning on-device image search engine using CLIP for embeddings and FAISS/USearch for efficient retrieval, ensuring full offline functionality and user privacy.

AI-powered smart glasses prototype using Python, Raspberry Pi, Gemini 2.0, and OpenCV to provide real-time voice feedback for detected objects and environmental cues.

An AI-powered tool that generates visually appealing summaries of YouTube videos, including notes, timestamps, and AI chat with the video content.

A real-time sign language to English converter using Python and TensorFlow, enabling seamless online conversations for Deaf individuals.

Developed a low latency speech-to-speech bot for more human-like conversation, reducing response time and improving natural interaction.

Created software to control car racing games using hand gestures, providing an innovative and immersive gaming experience without traditional controllers.

Developed an automatic attendance marking system for college using face recognition technology to streamline the attendance process.

Created a personal assistant for speech-to-speech conversation and small task completion using function calling capabilities.

Developed a system for detecting personal protective equipment (PPE) kits using computer vision techniques for safety compliance monitoring.