Job Description

Job Description

Job Description

Overview

We are seeking a high-agency Senior Computational Chemist to join PHIN. Your responsibility will be to constantly improve the architectural backbone of our materials discovery pipeline, bridging the gap between high-fidelity quantum chemistry and scalable machine learning. This role isn't just about running calculations; it’s about building the "plumbing" that allows us to explore the vast chemical space of materials at unprecedented speeds.

• Pipeline Automation: Design and implement automated workflows for DFT and Quantum Chemistry (QC) calculations to generate high-quality training data.

• ML Potential Development: Develop and deploy Machine Learning Potentials to enable large-scale simulations of industrially relevant systems and porous materials.

• Generative Modeling: Utilize generative models to bridge data generator and machine learning training length scales.

• Active Learning: Improve our active learning loops to minimize manual intervention and maximize the informational value of every CPU/GPU hour spent.

• Experience: 5–8 years of experience in computational chemistry, materials science, or a related field, with a proven track record of studying catalysts, polymers and MOFs.

• Generalist Capability: A "Swiss Army Knife" mentality—equally comfortable debugging a Python script, optimizing a basis set, or configuring a cloud cluster.

• HPC Familiarity: Extensive experience with SLURM and containerization (Docker/Singularity) for distributed computing

• Tool Agnostic: Skilled in multiple software packages (e.g., QE, VASP, CP2K, ORCA, Psi4) and able to pick up new tools rapidly based on the problem at hand.

• AI-Native: Deep intuition for how machine learning can accelerate traditional physics-based simulations, specifically regarding generative chemistry.

• ML Collaboration: Ability to work closely with data scientists to translate chemical problems into machine-learnable frameworks.

Please provide concise answers to the following questions as part of your application:

1. Automation: Describe a time you had to build a custom data pipeline or automation tool because existing software (e.g., QE, VASP, Gaussian) didn't support your specific workflow. What was the biggest bottleneck?

2. ML/Physics Trade-offs: When developing a Machine Learning Interatomic Potential (MLIP) for a complex system like a MOF, how do you decide when to prioritize more DFT training data versus increasing model complexity?

3. Generative Modeling: What is your experience in developing and using generative materials models?

4. Tool Agnosticism: Suppose we need to switch our primary simulation engine to a tool you’ve never used before (e.g., switching from CP2K to ORCA). Walk us through your process for becoming proficient and validating your results in this new environment within two weeks.

5. Scaling & Infrastructure: Describe your experience managing high-throughput calculations. How do you handle "job failures" at scale (e.g., 10,000+ simultaneous DFT runs) to ensure the integrity of the final dataset?

6. Values Alignment: Rate yourself 1-5 on the embodiment of the following values and give a 1 sentence rationale for each: People-first, Constant Improvement, Efficiency, High Quality, Visionary

Job Tags

Similar Jobs