ML Engineer, Physics-Informed Models & Real-Time Pipelines

Root Access, Inc.; New York (NY); On-site; USD 140,000 - 190,000; Full time

Job summary

Root Access, Inc. is a NYC-based frontier electronics startup building cutting-edge physics-informed AI. The team blends electrical, firmware, software, and ML expertise to push performance in real-time systems.

The role focuses on architecting physics foundation models, developing a robust ECAD data pipeline, and enabling seamless integration of graph neural networks with spatial physics engines, all while optimizing for GPU-based training and inference.

Qualifications

• Master’s or Ph.D. in a quantitative field with SciML focus.
• Deep learning framework experience with PyTorch or JAX.
• Hands-on SciML experience building/training PINNs, FNOs, etc.
• Strong mathematical depth in PDEs, vector calculus, autograd, and optimization.
• Experience manipulating spatial/geometric datasets with Python libraries.

Responsibilities

• Architect Physics Foundation Models: design and train deep learning models.
• Build the ECAD Data Pipeline: convert PCB data to continuous space.
• Multi-Modal Architecture Integration: connect GNNs/LLMs to spatial engines.
• Optimize for Real-Time Execution: optimize training/inference on GPU clusters.

Skills

PyTorch

JAX

SciML

PDEs

Automatic Differentiation

Numerical Optimization

Python

Data Pipelines

Education

Master’s or Ph.D. in Computer Science, Mathematics, EE, Physics

Tools

Open3D

Shapely

NumPy

SciPy

Job description

Root Access, Inc. is a NYC-based frontier electronics startup building cutting-edge physics-informed AI. The team blends electrical, firmware, software, and ML expertise to push performance in real-time systems.

The role focuses on architecting physics foundation models, developing a robust ECAD data pipeline, and enabling seamless integration of graph neural networks with spatial physics engines, all while optimizing for GPU-based training and inference.