Tayyab Aziz

My core interest lies in the foundational layer of Modern AI: Data. I specialize in engineering high-quality, synthetic datasets specifically designed for LLM Fine-Tuning and RAG (Retrieval-Augmented Generation) systems. I have developed cost-effective pipelines to generate large-scale training data by orchestrating top-tier APIs (Gemini 3, GPT-4o, Claude 3.5, and Perplexity). My focus is on reducing the cost of dataset creation while maximizing quality, diversity, and reasoning capabilities for custom model training.