GPU-Accelerated Edge AI/ML Engineer (Real-Time Wireless)

Booz Allen Hamilton; Annapolis (MD); On-site; USD 99,000 - 225,000; Full time

Job summary

Booz Allen Hamilton is seeking an AI/ML Engineer to join a team building a converged platform for real-time wireless signal processing and AI inference on GPU-accelerated hardware. You will develop, optimize, and deploy models alongside live network workloads in a lab environment.

The role covers data pipelines from telemetry to model training, GPU-optimized inference, and containerized deployment on Kubernetes with NVIDIA tools.

Qualifications

• 3+ years of experience training and evaluating ML models with PyTorch or TensorFlow, including exporting to ONNX
• 3+ years of Python software development with production-quality code
• Experience optimizing neural networks for real-time inference: latency/throughput, profiling, quantization or pruning
• GPU computing experience (CUDA/cuDNN) in Linux server environments
• Experience deploying ML models as services in Docker/Kubernetes and building data pipelines for time-series/telemetry
• Experience delivering in Agile cross-functional teams and solving complex problems with limited direction
• Ability to obtain a Secret clearance
• Bachelor's in STEM with 5+ years in AI/ML or Automation using Python, Ansible, Terraform, Kubernetes (or Master's 3+ years, or 8+ years in lieu of degree)

Responsibilities

• Design, train, and evaluate ML models for wireless network use cases using network telemetry and RF data
• Build data pipelines ingesting telemetry into reproducible training datasets
• Optimize models for real-time inference on GPU-accelerated edge servers with quantization/pruning and CUDA profiling
• Package and deploy models as containerized inference services on Kubernetes using Triton and MIG sharing
• Develop AI-driven network apps (xApps/rApps) with telemetry-based control decisions
• Instrument platform with MLOps tooling: experiment tracking, model registry, retraining triggers
• Benchmark GPU resource contention and document operating envelope for co-location
• Apply DoD AI assurance practices: provenance, data lineage, robustness testing
• Write technical docs and reproducibility guides in lab, not post-facto
• Collaborate with network/RF/cyber engineers to translate mission requirements into deployed models

Skills

ML frameworks (PyTorch / TensorFlow)

Python programming

GPU computing (CUDA / cuDNN)

Docker / Kubernetes

MLOps tooling

Security clearance (Secret)

DoD AI guidance

Education/Experience combo (STEM + AI/

C/C++ (systems programming)

NVIDIA Triton / TensorRT / ONNX

Education

B.S. in STEM + 5+ yrs AI/ML or Automation (Python, Ansible, Terraform, Kubernetes)

M.S. in STEM + 3+ yrs AI/ML or Automation (Python, Ansible, Terraform, Kubernetes)

8+ yrs AI/ML or Automation in lieu of degree

Tools

NVIDIA Triton Inference Server

NVIDIA GPU Operator

ONNX Runtime

CUDA profiling

Kubernetes

Docker

Job description

Booz Allen Hamilton is seeking an AI/ML Engineer to join a team building a converged platform for real-time wireless signal processing and AI inference on GPU-accelerated hardware. You will develop, optimize, and deploy models alongside live network workloads in a lab environment.

The role covers data pipelines from telemetry to model training, GPU-optimized inference, and containerized deployment on Kubernetes with NVIDIA tools.