AI Engineer / LLM Systems
I build applied AI systems across Agentic RAG, vector search, real-time data pipelines, and machine learning. My work combines practical deployment, data engineering, and LLM workflows for production-oriented problems.
RAG, LLM deployment, streaming data, forecasting
Based in Ho Chi Minh City, Vietnam. Currently studying Information Technology at the University of Information Technology, VNU-HCM, and working on applied AI systems.
Selected Work
Built a real-time cryptocurrency price pipeline using the Coinbase API, Kafka, and Spark. Stored high-frequency data in Cassandra, monitored services with Grafana, and trained an LSTM forecasting model that achieved R2 = 0.9677.
Developed a system that detects online deals and notifies users in real time. Fine-tuned LLaMA 3.1 8B for price prediction with Top-K decoding and weighted averaging, outperforming LightGBM, CatBoost, and Random Forest baselines.
Built a FastAPI RESTful service for an Agentic RAG system handling complex invoice queries. Combined semantic vector search, structured SQL queries, and OpenSearch entity resolution for a bilingual English/Japanese workflow.
AI Engineer Intern project at TGL Solution.
Capabilities
Background
Built an Agentic RAG service with FastAPI, LangGraph, Gemini, Qdrant, OpenSearch, SQL, and Docker. Focused on complex invoice queries and bilingual English/Japanese information retrieval.
Ranked Top 6 out of 617 teams by building a Select-Patch-Verify-Choose pipeline with open-weight LLMs via vLLM to automatically fix real GitHub issues.
Bachelor of Information Technology student with GPA 3.3/4.0. Certifications: IELTS 5.5 and TOEIC 725.
Contact
Reach out by email, phone, GitHub, or LinkedIn. My resume is available as a PDF download.