At SafiriAI, I designed and implemented data pipelines for logistics records to support AI model training and analytics. I also automated feature engineering and model inference workflows, reducing manual processing effort by approximately 70%.
As an Artificial Intelligence Intern at TPS Thanh Phong, I automated document processing, summarization, and content generation tasks, reducing estimated manual workload by 30–40%. I also created reusable prompt templates and supported practical AI training for staff of the People’s Committee of Binh Hung Commune.
For Sophia, my AI assistant for Marxist-Leninist philosophy, I developed a multilingual RAG system that provides evidence-based explanations with source citations. Its pipeline covers document parsing, semantic chunking, hybrid retrieval, reranking, and answer generation.
I also developed TiepLM, a NotebookLM-inspired assistant for deep-learning courses, and a multilingual text analysis tool for summarization and sentiment classification. My work draws on Python, FastAPI, PostgreSQL, React, and NLP.
