Heyuan LIU

Heyuan LIU

| École Polytechnique (l'X) |
| EPFL |
| Researcher |
| AI Engineer |



Looking For Ph.D. position in 2025 Fall and Research Internship From 2025 April



Energy System, Optimization, Machine Learning, Industrial System Modeling, Process Control

Download My Resume

About Me

Currently I just finished working as a Research Intern in EPFL IPESE Lab and Exchange Student in EPFL at Master level

I am a M2 AiViC (Artificial Intelligence and Advanced Visual Computing) student at École Polytechnique(Estimated graduate in October 2025)

My research is foucsing on bridging the advanced AI & CS techniques AND industrial Process & Energy System



Looking for: Internship from April to October 2025 and PhD positions in 2025 Fall

Ongoing Projects

Luggage Detection in 3D with IDEMIA

AI for Wild fire and extreme weather forecast with Prof. Qiuhua Huang in Colorado Mines

Research Interest

PINN and constraints programming

Machine Learning for Optimization in Large-Scale Industrial System

Learning Techniques for Process Control

Next generation of Decision-Making System with Data-driven Method in Industrial Process

Apply AI in energy and industrial process optimization & Reduce the Energy and Resource consumption in AI training Procedure

Education

Ecole Polytechnique Logo
École Polytechnique
Master of Science and Technology in Artificial Intelligence and Advanced Visual Computing
Coursework: Machine and Deep Learning, Image Analysis and Computer Vision, Computer Animation
Scholarship: SEMG scholarship
GPA: 3.67/4.0
Supervisor in M1: Prof. Jesse Read
Supervisor in M2: TBC (ETA March. 2025)

EPFL Logo
EPFL
Master Student in IPESE lab
Coursework: Modeling, optimization, design, and analysis of integrated energy systems
Supervisor: Prof. François Maréchal and Mr. Yi Zhao

Macau University of Science and Technology Logo
Macau University of Science and Technology
Bachelor of Science (Software Engineering)
Coursework: OOP in Java, Algorithms, Mobile Application Development, Data Science, Artificial Intelligence, Data Structures
Graduated with an Honor Degree
Supervisor: Prof. Rubing Huang

Work Experiences

IPESE Logo
EPFL IPESE lab
Research Intern
Mar. 2024 to Present
Applying advanced clustering and dimensionality reduction algorithms to preprocess datasets.
Designing ML models to train datasets and generate optimal solutions.
Analyzing results and providing recommendations with a LLM decision-making assistant.

RB Logo
Roland Berger
Feb. 2024 to Apr. 2024
Chatbot Engineer (PTA intern)
Developed Chatbot for car sales using coze and dify, created and deployed specialized agents: Sales, After-Sales, Service.
Ensured accurate responses and avoided chatbot hallucinations.

VW Logo
Volkswagen-Mobvoi
June 2022 to Aug. 2022
Software Quality Intern
Improved quality system from version 2.0 to 3.0, solved 36 software quality problems in Volkswagen ID6 and Audi A6 projects, ensured timely delivery of projects during COVID-19.

Projects

Project GIF

Identify optimal configuration with a machine learning method in multi-criteria decision analysis
Research Project @ EPFL IPESE lab
Mar. 2024 to Sep. 2024
Link: GitHub
The work in EPFL IPESE lab focuses on data-driven methods for decision-making in AI applications within Energy Systems and Industrial Processes. This involves techniques like Dimension Reduction, Clustering, Deep Learning, Reinforcement Learning, and the use of LLM Chatbots. The highlights are transform the original datasets into shapes and uses shape anaysis to extract more features

Project GIF

Navi-UAV
INF581 Project @ École Polytechnique
Jan. 2024 to Mar. 2024
Link: GitHub
Implemented Reinforcement Learning (DDPG) to enable UAVs to navigate through dynamic and uncertain environments efficiently.

Project GIF

Precision Collision Dynamics Simulation in Pinball
INF585 Project @ École Polytechnique
Jan. 2024 to Mar. 2024
Link: GitHub
Implemented the pinball simulations in aminations, hard and soft collisions are taken into considerations. (Iron pillar and Spring)

Project GIF

Real-Time AI for StarCraft
INF584A @ École Polytechnique
Jan. 2024 to Mar. 2024
Link: GitHub
Implemented real-time AI strategies for the game StarCraft based on the BWAPI, Selected Terran and beated 8/9 teams in the course, Ranking 1st in Terran, 2nd overall.

Project GIF

Extractive Summarization with Discourse Graphs
INF554 @ École Polytechnique
Oct. 2023 to Dec. 2023
Link: GitHub
Utilized GraphSAGE, Graph Convolutional Networks (GCN), Graph Attention Networks (GAT), and other graph-based models to enhance text summarization by analyzing conversation structures and important content.

Project GIF

Vtuber-Genshin
INF573 @ École Polytechnique
Oct. 2022 to Dec. 2023
Link: GitHub
Developed a real-time virtual character using MediaPipe and Unity, achieving high performance in head pose estimation and face mesh, ensuring stability and real-time capabilities.

Project GIF

A Dynamic Detection Approach for Oscillating Loss Problem in DNN based on AUTOTRAINER
Final Year Project @ MUST
Sep. 2022 - May 2023
Link: GitHub
Developed a dynamic bug detection system to identify and resolve oscillating loss issues in Deep Neural Networks, achieving a 50% reduction in time and computational resources required for bug detection, and received an “A+”.

Activity Experience

IEEE EI2 2024 conference SUB Reviewer
Sep. 2024 – Oct. 2024
IEEE Power and Energy Scociety, Will be held in Shenyang China, Dec. 2024

Zhejiang University SDG summer school
July 2023 – Aug. 2023
Focused on data visualization, including advanced research, practical training, and project reports.

Chinese Academy of Science summer research camp
July 2020 – Aug. 2020
Engaged in practical training and experiments on path planning for driverless cars, along with understanding current driverless technology developments.

Languages & Interests

Languages: English (Fluent), Chinese/Mandarin (Native), Deutsch (Basic), French (Beginner)

Interests: Basketball, Saxophone, Dragon boat

Special Awards

Project GIF

Issued by co-worker in École Polytechnique Ziyu WU
Translations: You are the backbone of our team. #Reliable teammate

Contact Me

Timeline of Projects