AI Scientist for Thermal Materials – Machine Learning & Property Validation _ CTO Office (Taipei)

<About the Role>
We are seeking an exceptional AI Scientist to pioneer the next generation of thermal storage materials using state-of-the-art artificial intelligence architectures.

Positioned at the intersection of artificial intelligence, thermodynamics, and materials science, you will establish the algorithmic foundation for discovering and validating thermal materials that are not only scientifically sound but also deployable in real-world energy systems.

Your primary mission is to leverage advanced AI models to understand how material structures and chemical interactions govern thermal performance and thermodynamic stability. By combining state-of-the-art architectures (such as Transformers, Diffusion models, or GNNs) with materials science, you will explore advanced materials—specifically Phase Change Materials (PCMs)—and translate fundamental physics into intelligent, predictive models.

Rather than performing isolated data science tasks, you will help build an AI-Based Materials Design Platform, providing reusable ML workflows, predictive design principles, and scalable algorithms that accelerate thermal materials innovation for next-generation heat recovery systems.

<Key Responsibilities>

AI-Driven Materials Discovery

• Design next-generation thermal materials using state-of-the-art AI architectures (Transformers, Diffusion, or GNNs).
• Explore vast chemical spaces to accelerate the discovery of materials with high thermal conductivity, energy density, and cycle stability.
• Establish AI-driven design strategies to replace traditional trial-and-error screening approaches.

Thermodynamic Property Validation

• Evaluate and validate thermodynamic behaviors using machine learning to ensure alignment with fundamental physics.
• Analyze predicted properties—such as phase stability, heat capacity, and thermal limits—to screen promising candidates.
• Provide algorithmic constraints to ensure models strictly adhere to thermodynamic laws.

Advanced Simulation & MLFF

• Construct and apply Machine Learning Force Fields (MLFF) to bridge quantum accuracy with large-scale simulation efficiency.
• Integrate AI algorithms with Molecular Dynamics (MD) simulations to evaluate atomic-level heat transfer and performance.
• Develop automated data pipelines to continuously train models on computational simulation results.

Cross-Disciplinary Translation

• Translate complex AI predictions into actionable insights for materials, mechanical, and experimental teams.
• Collaborate with domain experts to define physically meaningful descriptors and constraints for AI models.
• Partner with experimental and computational teams to validate AI predictions and refine ML methodologies.

<Required Technical & Domain Expertise>

1. Scientific Foundations

• Master’s or Ph.D. (preferred) in Materials Science, Mechanical Engineering, Chemical Engineering, Physics, Applied Mathematics, Computer Science, or related disciplines.
• Solid understanding of Materials Science principles, basic thermodynamics, and the underlying physics of material properties.

2. AI & Computational Modeling

• Extensive experience developing and implementing advanced AI research on state-of-the-art architectures, specifically Transformers, Diffusion models, or Graph Neural Networks (GNNs).
• Experience applying machine learning to physical sciences, particularly in materials discovery or property prediction.
• Proficiency in modern AI frameworks (e.g., PyTorch), Python, workflow automation, and handling large-scale scientific datasets.

3. Thermal Materials Knowledge Experience or familiarity with one or more of the following:

• Phase change materials (PCMs)
• Thermal storage materials
• High thermal conductivity materials