Ai Engineer Machine Learning

Thinh Nguyen

AI Engineer & Researcher NLP,

LLMs, and AI-Powered

Applications

Personal details

Thinh Nguyen

**********.****@*****.***

+316********

github.com/thinhrick0101

linkedin.com/in/thinh-

nguyenb31454215

Certificates

IBM AI Developer

Oct 2024

IBM

Machine Learning

Nov 2024

Stanford University

Deep Learning

Jan 2025

DeepLearning.AI

AI Agents Fundamentals

Feb 2025

Hugging Face

Profile

Innovative AI engineer with expertise in natural language processing, large language models, and AI-powered applications. Experienced in building and optimizing deep learning models, developing intelligent search systems, and deploying scalable AI solutions. Skilled in Python, Transformers, vector databases, and cloud-based AI integration. Passionate about leveraging AI to enhance user experiences and drive intelligent automation. Education

Bachelor's degree, Computer Science Aug 2022 - Aug 2025 Vrije Universiteit Amsterdam (VU Amsterdam), Amsterdam Current GPA: 8.2/10

Internships

NLP and Data Science Intern Mar 2025 - Jun 2025

ISODS, USA (remote)

• NLP pipeline development; semantic search implementations with FAISS and OpenAI.

• Integrated NLP solutions into client dashboards and enterprise workflows Projects

Vietnamese Restaurant Service Chatbot

Developed an AI-powered Vietnamese restaurant assistant using the LLaMA- 3.1-8B model with vector database integration.

Designed a recommendation system analyzing 30,000+ transactions via collaborative filtering and association rule mining. Built data pipelines to process sales data and generate personalized menu recommendations.

Implemented modular architecture with secure authentication and context- aware AI-driven responses.

Code Reviewer: AI-Powered Code Review & Refactoring Assistant Developed a Next.js web application powered by GPT-4 to automate comprehensive code review processes, providing real-time refactoring suggestions and bug detection.

Implemented a Retrieval-Augmented Generation (RAG) system using LangChain.js to integrate contextual insights from best-practice documentation.

Established persistent storage with MongoDB, allowing users to track submission history and iterative improvements.

Designed a responsive, user-friendly UI incorporating Monaco Editor for seamless code editing and comparative analysis of refactored suggestions. Fraudetect: Real-Time Fraud Detection System

Designed a comprehensive real-time fraud detection solution utilizing Kafka for event streaming, Apache Flink for real-time data validation and enrichment, and Spark for advanced feature engineering. Implemented MLflow for model lifecycle management, providing seamless deployment, monitoring, and tracking of predictive models. Optimized real-time risk scoring through Redis as a high-performance online feature store. Containerized the entire application using Docker for easy scalability and deployment on cloud infrastructure NeuroRate (BERT From Scratch for Sentiment Analysis) Built a BERT-based NLP model from scratch, achieving 80% accuracy on Amazon review sentiment analysis.

Engineered a custom Transformer encoder with WordPiece tokenization and memory-efficient training.

Optimized performance using mixed precision, gradient accumulation, and cosine learning rate scheduling.

Research Paper Assistant

Developed an AI-powered academic research tool integrating OpenAI API, FAISS vector search, and arXiv.

Engineered a hybrid search system combining semantic embeddings and keyword matching for optimal paper retrieval.

Built an automated summarization pipeline leveraging recursive LLM processing to extract key insights.

Designed a modular Python architecture with components for paper crawling, indexing, search, and analysis.

Created a Streamlit web interface for intelligent paper discovery via topic, title, or author search.

GenWeb: AI-Powered Website Generator

Built a full-stack AI website generator that converts text descriptions into deployable HTML, CSS, and JavaScript.

Integrated GPT-4o to generate structured, responsive code from natural language prompts.

Designed an intuitive Gradio UI for non-technical users to create professional websites.

Developed an automated one-click deployment pipeline to Vercel, streamlining website publishing.

Implemented intelligent image processing for media placement and code validation systems for best practices.

Utilized Python, JavaScript, REST APIs, and CI/CD in a modular, scalable architecture.

Contact this candidate