My Projects

VoiceGuru transforms learning with multilingual voice interaction (Kannada, Hindi, Tamil, English), an adaptive difficulty engine that classifies students as Beginner/Intermediate/Advanced, dynamic syllabus-aligned responses, a smart offline library with SQLite caching, gamified progress with badges and streaks, automated WhatsApp parent reports via Twilio, and a Vision Solver for homework images.

PromptGate gates LLM responses behind Algorand TestNet micropayments. Features include on-chain escrow (lock → release → payout), complexity-based auto model selection across Spark Lite / Balanced Core / Thorough Pro, publisher registration with pricing models, reputation & leaderboard tracking, a 3-stage autonomous pipeline (Research → Analysis → Summary), x402 payment challenge endpoints, and QR-based TestNet account verification.

An intelligent hospital management system that revolutionizes patient care through AI-driven appointment scheduling. The system prioritizes patients based on urgency and medical conditions, integrating real-time notifications via Twilio API. Features include smart queue management, doctor availability tracking, and automated patient communication to reduce wait times and improve healthcare delivery.

A comprehensive safety-first navigation system powered by machine learning algorithms. Analyzes real-time crime data, street lighting conditions, pedestrian traffic, and user feedback to provide the safest possible routes. Integrates with Google Maps API to offer multiple route options with safety ratings, making urban navigation secure for everyone, especially during night hours.

AI-based timetable generation system that uses constraint optimization to create practical schedules. Handles faculty and classroom conflicts through custom conflict-resolution logic, generating conflict-free timetables efficiently for academic institutions.

An advanced ML-powered platform that combats misinformation by detecting and classifying fake news articles with high accuracy. Utilizes NLP techniques, TensorFlow models, and scikit-learn algorithms to analyze text patterns, source credibility, and linguistic features. Includes a Streamlit dashboard for real-time news verification and content authenticity scoring.

A context-aware conversational assistant built with NLP and intent recognition. Supports multi-turn dialogue, persistent chat history via Firebase, and automated response flows. Built as an exploration of dialogue management before LLM APIs became mainstream.

Hackathon project that leverages computer vision and facial recognition to automatically detect and retrieve event photos. Upload your selfie and the system finds all event photos containing you using deep learning models, batch processing, and Firebase-backed storage.