AI Engineer in Training for Practical ML Solutions

DINH TRI TAI

Ho Chi Minh City +84-389-***-*** **********@*****.*** https://www.linkedin.com/in/dinh-tri-tai-625155405 https://github.com/tritai1 PERSONAL OBJECTIVE

I aim to become an AI Engineer who can design and deploy practical machine learning and LLM-based solutions. I am seeking opportunities to apply my skills in data analysis, model development, and real-world AI products while continuously improving my technical expertise and problem-solving ability.

EDUCATION

Posts and Telecommunications Institute of Technology (PTIT) 2022 – Present Bachelor of Engineering in Artificial Intlelligence (AI) English: Good at reading English documents and communicating at a basic level. Volunteer supporting digital transformation projects KEY PROJECTS & EXPERIENCE

Flight Price Prediction System (Repo) 2025

Machine Learning Pipeline

Collected and cleaned flight dataset from multiple fields (airline, route, departure time, duration, and stop count) Performed exploratory data analysis to identify key price drivers and seasonal patterns Applied feature engineering and preprocessing pipeline (categorical encoding, scaling, and train-test split) Trained and compared regression models including Decision Tree and Random Forest to select the best-performing model Integrated the trained Python model with a Node.js service to support real-time fare prediction from user input Student Lateness Prediction (Repo) 2025

Classification Model

Built a binary classification model to predict whether students would be late based on attendance and behavioral features Conducted end-to-end EDA to detect missing values, feature correlation, and class distribution imbalance Applied SMOTE to improve minority class representation and increase recall for late-student detection Engineered and selected meaningful features to improve model generalization on unseen data Deployed the inference pipeline as a real-time API for instant predictions and potential school monitoring workflows AI Article Retrieval Assistant (Repo) 2026

RAG / LLM Application

Designed end-to-end RAG pipeline: document loading, chunking, embedding, vector indexing, and retrieval Implemented chunk strategy with RecursiveCharacterTextSplitter (chunk size and overlap tuning) to improve context quality Built semantic retrieval with FAISS and sentence-transformers (all-MiniLM-L6-v2) for top-k relevant context selection Integrated retrieval output into LLM prompting flow to generate grounded, context-aware answers Optimized local embedding setup for stable and cost-efficient inference without external paid APIs Improved answer relevance by combining retrieval ranking and prompt context formatting for more accurate responses Selling Furniture — E-commerce Platform (Repo) 2025 Full-stack web application (Node.js + MongoDB + React) Developed backend APIs with Node.js, Express and Mongoose for products, users, carts, and orders Implemented authentication (Passport/JWT), file uploads (Multer) and email notifications Built responsive frontend with React, Vite and Tailwind; implemented product carousel and shopping cart UX Integrated Firebase and Google Maps services, and prepared the app for deployment (Vercel / hosting configs) Integrated Google Gemini (via @google/generative-ai) to enhance semantic product search and generate personalized recommendations

Implemented search flow: parse user query for intent, generate embeddings via Gemini, run vector similarity over product embeddings, then apply hybrid ranking with keyword signals. Used Gemini to re-rank results and generate short personalized recommendation snippets shown to users; added caching and rate- limiting to reduce latency and API costs.

Coordinated frontend/backend modules and improved user flows for browsing and checkout SKILLS

Programming: Python (Advanced), JavaScript (Intermediate), SQL (Intermediate), MongoDB Machine Learning & Data: Scikit-learn, Pandas, NumPy — modeling, feature engineering, evaluation Natural Language & LLMs: HuggingFace Transformers, RAG, FAISS, LangChain — embeddings, semantic search, prompt engineering, multi-hop retrieval, multi-LLM orchestration Tools & Deployment: Git, REST APIs, Firebase, Vercel, Jupyter Notebooks EXTRACURRICULAR

Digital Transformation Volunteer

AI Research Projects

Contact this candidate