Projects

A collection of things I've built — from AI systems to drones to backend tools.

#AI-Live Streamer

An AI-powered live streamer designed to answer science-related questions for children in the style of Albert Einstein, Elon Musk, and other science figures. Uses a custom-trained LLaMA model, custom AI voice TTS, EmoTalk for lip-sync, and a Unity 3D avatar. AI voice models on HuggingFace.

User input → Custom Trained LLaMA → TTS (custom voice model) → EmoTalk → Unity 3D Model

#TTS-RVC-API

Text-to-speech framework combining Coqui TTS with RVC (Retrieval-Based Voice Conversion) for voice cloning. Train a model with just 2–3 minutes of audio using Hubert for fast fine-tuning.

#OpenCV with DJI Drones

Use OpenCV with any DJI drone to access the real-time camera feed for live streaming, analysis, aerial photography, surveying, and drone-based monitoring.

#WhisCall

A framework for AI WhatsApp calls using Whisper, Coqui TTS, GPT-3.5 Turbo, Virtual Audio Cable, and the WhatsApp Desktop App.

#ESP-32 AI Drone

One of the most challenging projects — an ESP-32 based AI drone built on the crazyflie firmware, with modified IMU and barometer sensors and OpenCV integration for onboard computer vision.

#Raspi VTX System

A Raspberry Pi–based VTX system using packet injection, designed for FPV applications and managing drones, planes, and other UAVs. Demonstrated for military applications.

#Stratospheric Weather Data

A Teensy-based telemetry system using BMI270 (IMU), BMP180/BMP280 sensors, and ELRS for long-range communication. Collected temperature, altitude, pressure, and humidity data. Successfully tested to 20 km altitude. Demonstrated for military applications.