(Data-Driven Mechanism Research Track | 数据驱动机理研究方向)
We are hiring researchers under a unified role profile. This JD defines the core expertise areas we seek. Candidates may specialize in data-driven mechanism mining, multimodal analysis, or intelligent decision systems, and we will consider applicants with strengths in battery reliability or zero-carbon energy applications.
Candidates may demonstrate expertise in any subset of the listed domains. Applicants with strong depth in one area and willingness to collaborate across domains are encouraged to apply.
我们正在基于统一的岗位说明书招聘研究员。该 JD 定义了我们所需的核心能力领域。候选人可专注于数据驱动机理挖掘、多模态分析或智能决策系统，我们将考虑在电池可靠性或零碳能源应用方面具备优势的申请者。
候选人可在所列领域中的任意子方向展现专长。在某一方向具备深厚积累并愿意跨领域合作的申请者同样欢迎。
Responsibilities | 职责：
- Build and train AI foundation models to process heterogeneous data (time series, signals, simulations). Apply ML techniques (deep learning, XAI, causal inference) to uncover hidden correlations and mechanisms.
- Apply models to battery reliability (lifetime prediction, early warnings, thermal runaway diagnosis) and zero-carbon energy (power forecasting, load/price estimation, market trading decisions).
- Design intelligent decision-support systems to assist scientists/engineers in data analysis, hypothesis generation, experiment design, and decision-making. Explore LLM-based interfaces and autonomous AI agents.
- 构建并训练AI基础模型，处理异构数据（时序、信号、仿真等），利用深度学习、可解释性AI、因果推断等方法揭示隐藏相关性与机理。
- 将模型应用于电池可靠性（寿命预测、早期预警、热失控诊断）和零碳能源（功率预测、负荷与电价估计、电力市场交易决策）。
- 设计智能决策支持系统，辅助科学家和工程师进行数据分析、假设生成、实验设计与决策；探索基于LLM的人机交互界面与自主AI智能体。
Qualifications | 任职要求：
- PhD preferred (CS, Physics, Statistics, Automation, EE); outstanding Master’s considered.
- Passion for AI4S and applying data-driven methods to physical problems.
- Strong expertise in at least one of:1、Machine learning/deep learning (esp. time-series forecasting, Transformer/state-space models). 2、Multimodal data analysis (time series, signals, images).
- Excellent Python programming and algorithmic foundation; 3、familiarity with PyTorch or TensorFlow.
- 优先考虑计算机、物理、统计、自动化、电子工程等相关领域博士，优秀硕士亦可。
- 对AI4S充满热情，深刻理解数据驱动方法解决物理问题。
- 至少在以下方向之一具备专长：1、机器学习/深度学习（尤其是时间序列预测、Transformer/状态空间模型）。2、多模态数据分析（时序、信号、图像）。3、具备优秀的Python编程与算法功底，熟悉PyTorch或TensorFlow。
Preferred Qualifications | 优先条件：
- Proven experience applying data-driven methods to battery reliability or zero-carbon scenarios.
- Experience handling physical simulation (FEM, MD, CFD) or scientific instrument data (spectroscopy, microscopy, acoustic testing).
- Familiarity with battery failure experiments (ARC, DSC, TGA), feature extraction, and kinetic parameter inversion for model calibration and AI training.
- Internship/work experience in algorithm roles at major tech companies.
- Experience designing/developing AI agents, especially LLM-based.
- Publications in top-tier AI conferences or authoritative journals.
- 在电池可靠性或零碳场景中成功应用数据驱动方法的项目经验。
- 处理物理仿真（FEM, MD, CFD）或科学仪器（光谱、电镜、声学探伤）数据的经验。
- 熟悉电池失效实验（ARC, DSC, TGA），能够进行特征提取与动力学参数反演，用于模型标定与AI训练。
- 在大型科技公司算法岗位的实习或工作经验。
- 熟悉AI Agent（特别是基于LLM的Agent）的设计与开发。
- 在顶级AI会议或权威期刊发表过论文。
