Generative AI Research Engineer, ML Pipelines

Peraton Labs, Inc.; Red Bank (NJ); On-site; USD 146,000 - 234,000; Full time

Job summary

Peraton Labs is seeking an AI Research Engineer to lead the design and implementation of the machine learning core for generative-model research and development. The role focuses on model formulation, training/inference pipelines, and containerized deployment.

The ideal candidate has a research background in causal inference, probabilistic programming, or Bayesian methods and excels in a software engineering discipline.

Qualifications

• MS or PhD in CS/EE/Statistics/Applied Math or related field is required.
• 2+ years of applied ML research experience in causal inference, probabilistic programming, Bayesian methods, or learned optimization.
• Strong Python software engineering skills with production-quality PyTorch usage on GPUs.
• Experience turning research prototypes into maintainable, containerized code that runs unattended.
• Knowledge of multi-objective and constrained optimization applied to design spaces.

Responsibilities

• Design, implement, and train causal graph models and planning algorithms.
• Build ML infrastructure: training/inference pipelines, experiment tracking, dataset loaders.
• Package trained models for inference as containerized components with defined API.
• Contribute to synthetic data generation and simulation campaigns for training coverage.
• Present results at reviews and coordinate with academic subcontractors.

Skills

Python

PyTorch

ML research

Causal inference

Probabilistic programming

Bayesian methods

Multi-objective optimization

Experiment management

Reading research papers

US Citizenship

Education

MS or PhD in CS/EE/Statistics/Applied Math

Tools

GPU profiling

Container tooling

Job description

Peraton Labs is seeking an AI Research Engineer to lead the design and implementation of the machine learning core for generative-model research and development. The role focuses on model formulation, training/inference pipelines, and containerized deployment.

The ideal candidate has a research background in causal inference, probabilistic programming, or Bayesian methods and excels in a software engineering discipline.