Machine Learning Information Technology

HUNG NGUYEN (NGUYEN DUY HUNG)

PROJECTS

EDUCATION

Use case: Apply reforcement learning to flexible the time of the trafic light. Method: Testing q-learning and deep Q-learning on TSC by using SUMO simulator. Results: Reduce vehicle waiting time arverage 30%, deep Q-learning work better when trained for long periods of time. Beside that, Q-learning better at low trained. Link: https://github.com/21520894/reinforcement-learning-for-sumo Design automatic traffic lights

Oct 2024 - Dec 2024

ADDITIONAL INFORMATION

Soft skills: Research, Communication, Perseverance, Management, Problem-solving, Analytical thinking. Programming languages: Python, C#, C/C++, PHP, HTML/CSS, JavaScript, SQL, MySQL. Framework and lib: Tensorflow, Pytorch, Laravel, fastAPI, Pandas, Sklearn, Langchain. English: TOEIC L&R 725, S&W 280.

Predict next word using LSTM

Use case: Auto-complete sentance in message, in the product recommendation. Method: Using Trie for correct error word, deploy hybrid model include Seq2Seq LSTM for predicting next work and n-gram to choose number of words should be predicted. Results: Bleu score: 0.389, mean Bertscore: 0.67.

Link: https://github.com/21520894/auto-compleated-by-LSTM Mar 2024 - Jun 2024

Desktop Application

Use case: Predict commenter sentiment, evaluate student effectiveness, curriculum, recommendation system. Method: Developed a comprehensive desktop application to streamline student information management and enhance administrative efficiency.

Github: https://github.com/21520894/course-evaluation-based-on-comments Nov 2023 - Jan 2024

University of Information Technology - VNUHCM 09/2021 – 01/2025 GPA: 8.2/10.0

Major: Bachelor of Computer Science Current.

Relevant Courses: Data Structures & Algorithms, Object Oriented Programming, Operating System, Design & Analysis Algorithms, Data Mining, Machine Learning. Sep 2021 – Jan 2025

Phone: 081******* · Email: ***************@*****.*** · Github: github.com/21520894/ AI Engineering Intern

I am a highly motivated individual with a strong passion for artificial intelligence and machine learning, possessing a proactive attitude and a strategic mindset. My goal is to leverage my technical skills to develop innovative AI solutions that drive meaningful impact within organizations. Short and Mid-Term Objectives: Secure an internship in AI engineering, where I can apply my knowledge and contribute fully to the team’s success while continuously learning from experienced professionals. Long-Term Objectives: Advance to a full-time AI engineer position, ultimately aspiring to become a lead engineer contributing to cutting-edge projects that shape the future of technology. Object removal using image inpainting

Use case: Photoshop.

Method: Segmentation the object need to remove, use dialtion to enlarge the mask, avoid missing parts of the object. Convert model run time to cpu. Deploy and running model on SSH server. Minimize server-side resource requirements.

Result: Reach FID: 24.94, SSIM: 0.79, LPIPS: 0.18. Performance comparable to co-mod gan but reduces the size and time required to process an image.

Link: https://github.com/Duchai263/canvas_app

Mar 2023 - Jun 2023

Simple QA chatbot May 2025 - Present

Use case: Customer service bot, summary file.

Method: Simple QA chatbot for Retrieval data using LLM embedding, vector Store and Gemini API. Apply prompt engineering for limit the roles and result of chat bot. Using langchain framework for local store vector embedding and call gemini api.

Github: https://github.com/21520894/langchain

Contact this candidate