Projects
A collection of production systems, research projects, and technical explorations spanning AI, audio processing, and scalable infrastructure.
GPT 2 from Scratch
Implementation of GPT-2 language model from scratch, demonstrating text generation capabilities.
Music Style Transfer with RAVE
Real-time music style transfer using RAVE (Realtime Audio Variational autoEncoder) for creative audio manipulation.
Spleeter-HT-Demucs Stem Separation
Advanced audio stem separation using Spleeter and HT-Demucs models for high-quality vocal and instrumental isolation.
Evaluating and Improving Chain-of-Thought Reasoning
A framework for evaluating and enhancing Chain-of-Thought (CoT) reasoning in LLMs, including causal mediation analysis and faithfulness measurement.
Visual-guided prosody and emotional speech synthesis
Advanced text-to-speech system that uses visual cues to generate natural prosody and emotional expression in synthesized speech, creating more human-like and contextually appropriate voice output.
Automated TTS Dataset Generation
Automated workflow to generate TTS datasets using A.I. for efficient text-to-speech model training.
Flowent - World's First Cognitive-Science Approved Immersion Language Learning App
A language learning app that is grounded in cognitive science and uses the latest in AI to help you learn languages faster and more effectively.