Wireless and Edge AI/ML Engineer

Booz Allen Hamilton•Annapolis Junction, MD

9h•$99,000 - $225,000•Hybrid

About The Position

Booz Allen is seeking an AI/ML Engineer to help build a converged platform that runs real-time wireless signal processing and AI inference on the same GPU-accelerated hardware. The engineer will develop, optimize, and deploy models alongside live network workloads, where latency is measured in microseconds and computing resources are shared with the radio stack. The work covers the full lifecycle, from network-telemetry data pipelines and model training to GPU-optimized inference and containerized deployment on Kubernetes in a hands-on lab. The role involves applying machine learning to wireless network use cases, optimizing inference on edge servers, and developing AI-driven network applications. The engineer will instrument the platform with MLOps tools, benchmark resource contention between signal-processing and AI workloads, and apply security and assurance practices throughout the ML lifecycle. They will document work as it proceeds, collaborate with network, RF, and cybersecurity engineers, and brief results to government stakeholders. The role is hybrid and based in Annapolis Junction, Maryland. Applicants must be able to obtain a Secret clearance.

Requirements

At least 3 years of experience training and evaluating ML models using frameworks such as PyTorch or TensorFlow, including exporting models to deployment formats such as ONNX.
At least 3 years of Python software-development experience, including production-quality code, testing, and version-control practices.
Experience deploying ML models as services in Kubernetes.
Experience optimizing and benchmarking trained neural networks for inference, including latency and throughput measurement, profiling, quantization, or pruning.
Experience with GPU computing, including CUDA, cuDNN, or GPU-accelerated libraries, and NVIDIA driver and container tooling in a Linux server environment.
Experience building data pipelines for time-series or telemetry data, including feature engineering and dataset versioning.
Ability to work hands-on in a fast-paced lab environment with hard delivery deadlines.
Ability to produce clear, accurate technical documentation and brief results to technical and non-technical audiences.
Ability to obtain a Secret clearance.
A bachelor's degree in a STEM field and at least 5 years of experience in AI/ML or automation using Python, Ansible, Terraform, and Kubernetes; or a master's degree in a STEM field and at least 3 years of that experience; or at least 8 years of that experience in lieu of a degree.

Nice To Haves

Experience with RF theory, digital signal processing, or physical-layer algorithms such as channel estimation, beamforming, and modulation classification, including GNU Radio or MATLAB.
Experience with AI-RAN or AI-native wireless platforms, such as NVIDIA Aerial, GPU-accelerated Layer 1, or AI/ML-based RAN optimization.
Experience with O-RAN architectures, including Near-RT and Non-RT RAN Intelligent Controllers, xApp/rApp development, and E2 and A1 interfaces.
Experience using 5G/LTE network KPIs, performance counters, and telemetry, such as 3GPP performance-management data, as ML training inputs.
Experience deploying and optimizing large language models for inference using frameworks such as vLLM or NVIDIA TensorRT-LLM, including GPU memory optimization and latency and throughput benchmarking.
Experience with neural network architectures including transformers, graph neural networks, and reinforcement learning for control and scheduling problems.
Experience with C/C++ or other systems programming languages for performance-critical applications or GPU-accelerated computing.
Experience with MLOps platforms and tooling such as MLflow, Kubeflow, DVC, or equivalent.
Experience working in an Agile, cross-functional engineering team, resolving complex problems with limited direction, and developing automated tests for ML data pipelines, model validation, and performance regression testing.
Experience with DoD AI assurance and responsible AI guidance, and DoD STIG or RMF requirements for ML systems.

Responsibilities

Design, train, and evaluate ML models for wireless network applications, including traffic prediction, anomaly detection, interference classification, resource scheduling, and energy optimization, using network telemetry, KPIs, and lab-captured RF data.
Build pipelines to ingest, label, and version network-stack telemetry—including performance counters, logs, packet captures, and I/Q samples—into reproducible training datasets.
Optimize trained models for real-time inference on GPU-accelerated edge servers using techniques such as quantization, pruning, TensorRT or ONNX Runtime conversion, and CUDA profiling; benchmark latency and throughput against targets.
Package and deploy models as containerized inference services on Kubernetes using NVIDIA Triton Inference Server, the NVIDIA GPU Operator, and MIG or time-sliced GPU sharing alongside live network workloads.
Develop AI-driven network applications, including xApps and rApps, that consume telemetry and return control decisions through standardized interfaces; validate them end to end in the lab.
Instrument the platform with MLOps tooling for experiment tracking, model registry, automated retraining triggers, drift monitoring, and CI/CD for model deployment.
Benchmark GPU resource contention between signal-processing and AI-inference workloads, and characterize and document the operating envelope for co-location on a single platform.
Apply ML lifecycle security and assurance practices, including model provenance, dataset lineage, adversarial robustness testing, and alignment with DoD AI assurance guidance.
Write and maintain technical documentation, including model cards, training and deployment procedures, benchmark reports, and reproducibility guides, alongside lab work.
Collaborate with network, RF, and cybersecurity engineers to translate mission requirements into models for the deployed system, and brief results to government stakeholders.