Software Engineer, Perception

This is a fully remote position, open to applicants in Arizona.

📋 Description

• Create real-time detection, classification, and multi-object tracking solutions utilizing live sensor feeds.

• Integrate data from various sensors into a unified track picture.

• Design, train, and assess models to meet latency and accuracy targets for embedded deployment.

• Enhance inference performance on embedded hardware through techniques such as pruning, quantization, and TensorRT.

• Develop and oversee the dataset pipeline, which includes collection, labeling, curation, augmentation, synthetic data generation, and retraining using field data.

• Establish the evaluation framework and metrics to be assessed under operational conditions.

• Provide track state information (position, velocity, identity, uncertainty) to control software and the operator interface.

• Create backend services and APIs as needed.

• Collaborate with hardware engineers on the selection of cameras, optics, and sensors.

• Assist with integration and testing, including troubleshooting and resolving issues on target hardware.

⛳️ Requirements

• Proven experience in developing real-time computer vision and multi-object tracking systems.

• Experience in delivering detection or tracking systems to production, particularly under latency, small-object, and adversarial conditions.

• Familiarity with sensor fusion techniques.

• Background in backend development, integration, and systems software in addition to perception.

• Proficient in Python and C++ programming.

• Experience working with hardware and handling real-time data.

• Must be a U.S. citizen.

• Preferred: Experience with PyTorch, TensorRT, ONNX, DeepStream, or GStreamer.

• Preferred: Familiarity with thermal, infrared, or multispectral imaging technologies.

• Preferred: Experience in multi-modal sensor fusion involving EO/IR, radar, or acoustic sensors.

• Preferred: Knowledge of synthetic data generation and sim-to-real processes.

• Preferred: Experience deploying on NVIDIA Jetson platforms.

• Preferred: Familiarity with embedded Linux and device I/O (cameras, gimbals, sensors).

• Preferred: Knowledge of gRPC or WebSockets.

• Preferred: Experience in defense or other safety-critical computer vision applications.

• Preferred: Involvement in personal robotics or hardware projects.

🏝️ Benefits

• Fully remote position.