11/06/2026 updated
40 % partially available
Computational Chemist | Molecular Simulation & Scientific Software | CR Scientific
New York, USA
Only remote
Ph.D. in Chemical Engineering, University of California, BerkeleyAbout me
I develop molecular simulations and scientific software to connect physical models with practical research questions. Through CR Scientific, I support work in pharmaceuticals and energy, drawing on my research in statistical mechanics, molecular modeling, and high-performance computing.
Computational ChemistryNvidia CUDAMachine LearningMolecular ModellingHigh Performance ComputingNumerical Computing
Machine Learning & AI for Molecular Modeling
Expertise in developing graph neural network architectures (PyTorch, e3nn) for molecular property prediction, achieving state-of-the-art performance on standard benchmarks. Experience building end-to-end ML pipelines orchestrating multiple AI models including AlphaFold for automated multi-stage drug screening and property prediction.
High-Performance Computing & GPU-Accelerated Simulation
Proficiency in developing GPU-accelerated computing infrastructure in C++/CUDA for large-scale Monte Carlo and physics-based simulations, delivering 10-100x speedups. Skilled in parallel algorithm engineering using MPI for distributed multi-node scaling.
Scientific Software Development & Open Source Engineering
Strong background in building production-ready scientific software, including a GPU-accelerated simulation engine (cav-hoomd), a JIT-compiled numerical computing library for quantum dynamics (SQD-JIT), and open-source Python packages for large-scale numerical linear algebra (pyglasstools).
Molecular Simulation Tools
Hands-on experience with leading molecular simulation platforms including HOOMD-Blue, LAMMPS, OpenMM, Desmond, and ASE for large-scale physics-based and classical molecular dynamics simulations.
AI Engineering & Agentic Workflows
Experience prototyping agentic AI workflows with retrieval-augmented generation (RAG) using LangChain and MCP to coordinate multi-model pipelines for automated design-evaluate loops in materials discovery.
Electronic Structure & Finite Element Analysis
Proficiency in electronic structure tools (PySCF, ORCA, CP2K) and finite element analysis software (COMSOL, PyBAMM, PETSc/SLEPc, FEniCSx) for computational chemistry and engineering applications.
Data Pipeline Development & Automation
Deployed automated high-throughput data-generation pipelines across 10,000+ molecular compositions, producing 1M+ labeled records for downstream model training and evaluation. Co-led dataset development for Meta FAIRs Open Molecules 2025 initiative contributing to a ~10-petabyte dataset of 100M+ molecular structures.
Programming & HPC Environments
Skilled in C/C++, Python, Mathematica, MATLAB, CMake/Make, Bash, and Git. Experienced with HPC environments including OpenMP, OpenMPI, CUDA, SGE/PSB, and SLURM for large-scale scientific computing.
Differentiable & Optimization-Based Modeling
Developed differentiable predictive models using JAX for client optimization problems in energy systems, integrating domain constraints with gradient-based parameter estimation.
Markup, Graphics & Visualization
Proficient in LaTeX, Markdown, RMarkdown, Hugo, Tikz, gnuplot, Matplotlib, Inkscape, Adobe Illustrator, Avogadro, and Ovito for scientific communication and data visualization.
Expertise in developing graph neural network architectures (PyTorch, e3nn) for molecular property prediction, achieving state-of-the-art performance on standard benchmarks. Experience building end-to-end ML pipelines orchestrating multiple AI models including AlphaFold for automated multi-stage drug screening and property prediction.
High-Performance Computing & GPU-Accelerated Simulation
Proficiency in developing GPU-accelerated computing infrastructure in C++/CUDA for large-scale Monte Carlo and physics-based simulations, delivering 10-100x speedups. Skilled in parallel algorithm engineering using MPI for distributed multi-node scaling.
Scientific Software Development & Open Source Engineering
Strong background in building production-ready scientific software, including a GPU-accelerated simulation engine (cav-hoomd), a JIT-compiled numerical computing library for quantum dynamics (SQD-JIT), and open-source Python packages for large-scale numerical linear algebra (pyglasstools).
Molecular Simulation Tools
Hands-on experience with leading molecular simulation platforms including HOOMD-Blue, LAMMPS, OpenMM, Desmond, and ASE for large-scale physics-based and classical molecular dynamics simulations.
AI Engineering & Agentic Workflows
Experience prototyping agentic AI workflows with retrieval-augmented generation (RAG) using LangChain and MCP to coordinate multi-model pipelines for automated design-evaluate loops in materials discovery.
Electronic Structure & Finite Element Analysis
Proficiency in electronic structure tools (PySCF, ORCA, CP2K) and finite element analysis software (COMSOL, PyBAMM, PETSc/SLEPc, FEniCSx) for computational chemistry and engineering applications.
Data Pipeline Development & Automation
Deployed automated high-throughput data-generation pipelines across 10,000+ molecular compositions, producing 1M+ labeled records for downstream model training and evaluation. Co-led dataset development for Meta FAIRs Open Molecules 2025 initiative contributing to a ~10-petabyte dataset of 100M+ molecular structures.
Programming & HPC Environments
Skilled in C/C++, Python, Mathematica, MATLAB, CMake/Make, Bash, and Git. Experienced with HPC environments including OpenMP, OpenMPI, CUDA, SGE/PSB, and SLURM for large-scale scientific computing.
Differentiable & Optimization-Based Modeling
Developed differentiable predictive models using JAX for client optimization problems in energy systems, integrating domain constraints with gradient-based parameter estimation.
Markup, Graphics & Visualization
Proficient in LaTeX, Markdown, RMarkdown, Hugo, Tikz, gnuplot, Matplotlib, Inkscape, Adobe Illustrator, Avogadro, and Ovito for scientific communication and data visualization.
Experience
Professional experience5 - 10 years
Freelancer since2025
Languages
EnglishFluent
Project history
Founded a solo consulting practice delivering AI and scientific computing projects for clients in energy and pharmaceuticals. Conducted due diligence on client IP and R&D proposals. Built end-to-end ML pipelines for automated multi-stage drug screening, developed differentiable predictive models (JAX) for energy system optimization, and prototyped agentic AI workflows with RAG using LangChain and MCP.
Simons Postdoctoral Fellow leading independent research in AI-driven molecular modeling and high-performance simulation. Co-led dataset development for Meta FAIRs Open Molecules 2025 initiative. Deployed automated high-throughput data-generation pipelines, developed graph neural network architectures for molecular property prediction, built a GPU-accelerated simulation engine (cav-hoomd) in C++/CUDA, and built SQD-JIT, a JIT-compiled numerical computing library for quantum dynamics.
Developed GPU-accelerated computing infrastructure in C++/CUDA for large-scale Monte Carlo and physics-based simulations. Engineered a distributed parallel algorithm (MPI) for low-probability event estimation. Trained deep learning models (PyTorch) on high-dimensional simulation trajectories. Built pyglasstools, an open-source Python package for large-scale numerical linear algebra with distributed compute backends.