Postdoc: AI-Driven Atomistic Simulations for Materials Design

Laplace Expert Ltd; Maryland; On-site; USD 60,000 - 80,000; Full time

Job summary

Johns Hopkins University in Maryland invites applications for a Postdoctoral Fellow in atomistic simulations and AI for materials design. The role emphasizes DFT, machine-learned force fields, graph neural networks and large language models to accelerate materials discovery.

The AtomGPTLab will build a computational platform integrating first-principles methods, molecular simulations, GNNs and LLMs to enable faster experimental design, data integration, open-source contributions, mentoring and

Qualifications

• Ph.D. in Materials Science, Physics, Chemistry, Chemical Engineering, Computer Science or related field is required.
• Experience in at least one of density functional theory, machine-learned force fields, graph neural networks or large language models.

Responsibilities

• Lead high-throughput DFT studies and develop machine-learned force fields.
• Develop graph neural networks and large language models for materials design.
• Contribute open-source software, mentor students, and prepare manuscripts.
• Collaborate across disciplines to accelerate materials discovery.

Skills

Interdisciplinary collaboration

Mentoring

Manuscript writing

Grant proposal contribution

Scientific communication

Education

Ph.D. in Materials Science/Physics/Chemistry/Chemical Engineering/Computer Science or related field

Tools

Density functional theory

Machine-learned force fields

Graph neural networks

Large language models

VASP

Quantum ESPRESSO

GPAW

ALIGNN-FF

ALIGNN

HuggingFace

OpenAI

AtomGPT

JARVIS-Tools

ASE

High-performance computing

Python

Git

Conda packaging

Surrogate modeling

Graph neural networks

Large language models

Job description

Johns Hopkins University in Maryland invites applications for a Postdoctoral Fellow in atomistic simulations and AI for materials design. The role emphasizes DFT, machine-learned force fields, graph neural networks and large language models to accelerate materials discovery.

The AtomGPTLab will build a computational platform integrating first-principles methods, molecular simulations, GNNs and LLMs to enable faster experimental design, data integration, open-source contributions, mentoring and