ABOUT US:
At Quartermaster AI, we believe the ocean should be a safe and sustainably managed resource for all. By leveraging cutting-edge AI and robotics, we unlock capabilities that were only recently impossible. Our distributed open-ocean systems enable every vessel to sense, compute, and communicate, enhancing maritime domain awareness for those who need it most.
JOB DESCRIPTION:
Quartermaster AI is seeking a Senior AI/ML Engineer with an emphasis in RF analysis to develop and deploy machine learning systems that utilize RF data for real-time maritime intelligence.
You’ll work in a small team of experienced engineers to build detection, classification, and tagging models that help provide contextual understanding of vessel activity based on observed RF signatures.
KEY RESPONSIBILITIES:
- Design, train, and deploy machine learning models for RF signal detection, classification, and vessel activity tracking.
- Build and maintain dataset curation pipelines, including AIS-correlated ground truth labeling, synthetic RF data generation, and augmentation strategies for class-imbalanced maritime environments.
- Build the interface between DSP feature outputs and model inputs by defining pre-processing, normalization, and feature extraction requirements in coordination with the DSP engineer.
- Develop model evaluation frameworks and benchmarking harnesses; define quantitative performance criteria and drive iterative improvement against them.
- Optimize models and inference workflows for deployment on edge compute hardware.
- Document model architecture, training methodology, dataset provenance, and validation results.
QUALIFICATIONS (PREFERRED):
- Master's or PhD in Machine Learning, Signal Processing, or a closely related field — or equivalent demonstrated experience.
- 5+ years building and deploying ML systems with a focus on RF or signals data.
- Proficiency in Python and deep learning frameworks; familiarity with RF-native tooling such as Torchsig is a strong plus.
- Strong understanding of signal alignment, temporal synchronization, and feature extraction from IQ and spectral data.
- Proven ability to ship production models, not just research prototypes.
- Experience in maritime, aerospace, or operationally demanding spectral environments.
- Experience building labeled RF datasets from ground truth sources.
- Familiarity with edge inference constraints and optimization techniques (quantization, pruning, model distillation).
- Active Secret clearance or demonstrated ability to obtain one.