Projects

Truth Guard

Developed a system to detect fake content across audio, video, images, and text using CNN + Librosa (audio), CNN + LSTM (video), ResNet-50 (images), and Passive-Aggressive Classifier (text).
Built a full-stack mobile application with Django backend and Flutter frontend, connected via REST APIs for real-time detection and verification.
Engineered the system to efficiently analyze and classify fake news and media, demonstrating end-to-end AI/ML integration in a practical app.

Developed an app supporting 22 Indian languages, enabling real-time cross-language voice and text communication.
Implemented automatic speech-to-text translation and text-to-speech output, allowing seamless interaction between users speaking different languages.
Integrated Google Translate API and Google Cloud services, containerized with Docker, and deployed for scalable, high-performance operations.
Built with Java and Android Studio, delivering a robust, user-friendly platform that bridges linguistic barriers across India.

Developed an advanced system to provide real-time forecasts of thunderstorms and gale-force winds, enhancing aviation safety.
Combined Numerical Weather Prediction (NWP) models with logistic regression for historical data analysis and deep neural networks to capture complex patterns in large datasets.
Created 3D radar visualizations using Unity, enabling pilots and aviation personnel to intuitively assess storm severity, movement, and impact.
Delivered a multi-faceted AI/ML solution that improves meteorological forecasting for critical aviation decision-making.