Aakash Dwivedi

I am an AI/ML-focused software engineer with hands-on experience building real-time speech systems and NLP applications. At Samsung R&D I developed a real-time language-based audio denoising pipeline and implemented efficient language identification using quantized whisper models for low-latency production use.

I design end-to-end systems and full-stack products — from ML models to user-facing dashboards. I built STUTOR, a peer-to-peer learning platform with WebRTC video, and an NLP-based programming assessment tool that uses Llama 3 fine-tuned with LoRA and T5 for distractor generation and personalized recommendations.

I combine strong software engineering skills in Python, JavaScript, and backend frameworks with practical ML techniques (LLMs, fine-tuning, QLoRA) to deliver performant, production-ready solutions. I pursue roles where I can drive applied research into deployed products and continue improving real-time ML systems.