I am Zhu Enze(朱恩泽), a 2025 undergraduate student in the School of Automation at Xi'an Jiaotong University(西安交通大学) and a member of the 2023 class of the Young Gifted Program (YGP, 少年班).
My focus lies at the intersection of intelligent systems and software engineering, with a proven ability to solve practical problems through independent development and project implementation. My interests span machine learning, embedded systems, and full-stack application development.
Each card highlights one stage of training, from foundational preparation to formal degree study.
Bachelor of Engineering in Automation
Full undergraduate study in automation, intelligent systems, and engineering fundamentals.
Currently pursuing formal undergraduate training with an emphasis on control, software, and interdisciplinary engineering practice.
Young Gifted Program (Preparatory Phase)
Accelerated preparatory training bridging advanced pre-college coursework and early undergraduate practice.
Completed prerequisite coursework ahead of schedule, explored AI and software projects independently, maintained a technical blog, and built an early record in competitions and software copyrights.
A reverse-chronological timeline of research, engineering, and independent builds.
A serverless-friendly SVG card generator for Bilibili user profiles. Implements a clean API that fetches user stats via Bilibili web endpoints, transforms data with sanitization and formatting, then renders responsive vector cards using template-based SVG composition. Features a decoupled theme-color system, automated text wrapping for bios/titles, parallel async fetching with built-in caching, and HTML escaping for XSS prevention. Deployed on Vercel as serverless functions with a fallback Python script for local testing.
Tech Stack: JavaScript, Python, HTML/CSS, SVG templates, Vercel
Deployed and maintained a home server cluster on Orange Pi, stably running various services including Ollama (local large language model service), Nextcloud (private cloud storage), WordPress (website management), etc. Gained practical experience in Linux system administration, containerized application deployment, and intranet penetration techniques.
Tech Stack: Linux, Docker, Orange Pi
Responsible for the multimodal data preprocessing pipeline, specifically handling the download, cleaning, and spatial interpolation of ERA5 reanalysis meteorological data to provide structured input for subsequent CNN and large language model training, and made innovations based on the YingLong Model.
Tech Stack: Python, PaddlePaddle
Outcome: The project won a provincial third prize in the AIC competition.
To overcome the limitation of lacking a dedicated signaling server, I established a signaling channel using Firebase Realtime Database. Combined with WebRTC technology and public STUN servers, this system enabled direct peer-to-peer file transfer between browsers.
Tech Stack: JavaScript (WebRTC API), Firebase
Designed and implemented a lightweight remote control solution based on web and local bridge programs. Utilized Firebase for real-time signaling synchronization and employed a local Node.js script to parse commands and simulate keyboard events, enabling seamless control of PC-side PPT presentations from a mobile device.
Tech Stack: JavaScript, Node.js, Firebase Realtime Database
Designed, developed from scratch, and continuously maintain a static blog system for sharing technical notes and project summaries. The system achieves zero server backend deployment, with all code open-sourced. The site has obtained a software copyright.
Tech Stack: JavaScript, HTML/CSS, Jekyll
Contributed to partial code development for a weather forecasting model based on a U-Transformer structure. The design integrates strengths from both Pangu and Fuxi large weather models, implemented using the domestic PaddlePaddle deep learning framework to support efficient spatial-temporal prediction.
Tech Stack: Python, PaddlePaddle, NumPy
Outcome: Functional prototype for multi-step meteorological variable forecasting.
Responsible for dataset downloading, preprocessing, and final reporting. The innovation lies in training the CNN from scratch without freezing earlier layers, along with introducing improvements in pooling strategies and cosine annealing scheduling to enhance convergence and accuracy.
Tech Stack: Python, PyTorch, OpenCV
Outcome: Course project submitted with full pipeline documentation and performance analysis.
Selected competition results, scholarships, and distinctions.
Provincial Third Prize
Responsible for downloading, preprocessing, cleaning, and spatial interpolation of multimodal meteorological data.
National First Prize (First Place in the Northwestern Division)
Responsible for animation simulation and video production of core experimental phenomena.
Third class scholarship at the school level
Top 20% in comprehensive testing ranking.
Honorable Mention
Responsible for mathematical model construction, paper writing, and data visualization.
Core tools and strengths across software, machine learning, and embedded practice.
A compact milestone stream with recent academic, project, and competition updates.
Our team won the National First Prize in the China Undergraduate Physics Experiment Competition.
I received the school-level scholarship from Xi'an Jiaotong University.
Our team received the Honorable Mention in the Mathematical Contest in Modeling / Interdisciplinary Contest in Modeling.
The "Quark Blog" project was officially registered for software copyright.
My personal technical blog "Quark Blog" (lsqkk.github.io) went live.
I obtained the software copyright for the "Quark Autonomy Control" project.
This academic homepage template is a deeply customized fork of the open-source Minimal Light theme. The live repository board below is loaded from the GitHub API.
Loading repository metadata from GitHub...