Embedded Edge AI Engineer for Real-Time Aerial Autonomy

Society of Defense Financial Management; Charlotte (NC); On-site; USD 120,000 - 180,000; Full time

Job summary

PowerusPowerus is seeking an Edge AI Engineer to deploy and validate ML models on embedded hardware for autonomous aerial systems. The role covers model conversion, quantization, and on-target inference pipelines, collaborating with autonomy and flight software teams.

You'll optimize performance with TensorRT/ONNX/CUDA, profile on-device metrics, and ensure robust deployment through HIL testing and documentation.

Qualifications

• Must have a Bachelor's degree in a related field or equivalent experience.
• Experience deploying ML models to embedded/edge hardware with C++/Python.
• Proficient in TensorRT, ONNX, and CUDA for optimization.
• Hands-on experience with PyTorch or TensorFlow and embedded Linux.
• U.S. Person status required for ITAR/EAR access.

Responsibilities

• Convert, optimize, and deploy trained models to embedded GPUs/accelerators (NVIDIA Jetson).
• Apply quantization (INT8/FP16) and pruning to meet latency and power targets.
• Build inference pipelines using TensorRT, ONNX Runtime, CUDA.
• Integrate inference with camera/sensor/flight software in C++/Python.
• Profile latency, throughput, memory, power, and thermal behavior on target hardware.
• Define acceptance criteria for deployed models and maintain QA tooling.
• Document deployment configs and interface specs; support HIL/flight tests.

Skills

C++

Python

TensorRT

ONNX

CUDA

PyTorch

TensorFlow

Embedded Linux

Education

Bachelor's degree in computer engineering / CS / EE

Tools

NVIDIA Jetson

JetPack

Job description

PowerusPowerus is seeking an Edge AI Engineer to deploy and validate ML models on embedded hardware for autonomous aerial systems. The role covers model conversion, quantization, and on-target inference pipelines, collaborating with autonomy and flight software teams.

You'll optimize performance with TensorRT/ONNX/CUDA, profile on-device metrics, and ensure robust deployment through HIL testing and documentation.