> boot sequence initiated
> operator ............... ZHANG YUHENG
> role ................... Data & Algorithm Engineer
> focus .................. ML systems / CV / Time Series / LLM knowledge extraction
> stack .................. Python / C++ / CUDA / Java / SQL
> status ................. OPEN TO INTERESTING WORK
|
I build production-facing AI systems with a research mindset and an engineer's bias for execution. My work sits around machine learning, deep learning, computer vision, time series forecasting, high-performance C++ acceleration, and LLM-enhanced workflows. 目前主要方向是机器学习系统、时间序列预测、计算机视觉与大语言模型知识提取。我希望把研究能力和工程落地能力真正结合起来，做有实际业务价值的智能系统。
|
name: Zhang Yuheng
base: Chengdu, China
role: Data & Algorithm Engineer
languages:
- Python
- C++
- CUDA C
- Java
- SQL
interests:
- Machine Learning
- Time Series Forecasting
- Computer Vision
- Distributed Systems
- Research Engineering
|
- Predictive maintenance and KPI optimization for pharmaceutical equipment.
- Freeze dryer diagnostics, fluid bed process prediction and tablet press quality control.
- Data-driven monitoring combined with operational decision support.
- Built a TCN-based severe convection forecasting pipeline.
- Used C++ parallel preprocessing to improve rolling prediction efficiency.
- Designed and developed a modular C++ algorithm platform for AR scenarios.
- Covered algorithm interfaces, GPU scheduling and edge inference migration.
- Working on reliable knowledge extraction from LLMs for time series forecasting.
- Focusing on structured knowledge, reliability scoring and compact knowledge encoding.
