Embedded AI Engineer - Edge AI for Public Safety

About the role:

We’re looking for an Embedded AI Engineer to lead the development and deployment of deep learning models on IoT and edge computing platforms. You’ll work on converting and optimizing computer vision models—such as CNNs, YOLO, RNNs, and OCR—for execution on high-efficiency hardware accelerators like Hailo, NVIDIA Jetson, Intel Movidius, or ARM-based NPUs.

This role is central to scaling client's next-generation Edge AI platform, enabling real-time safety decisions and unlocking new capabilities for school bus fleets, student safety, and public transportation environments.

About Client:

The client is transforming how communities protect students and enforce traffic safety—through intelligent, real-time video analytics and automated enforcement. As part of this mission, we are advancing the shift from cloud-dependent AI to intelligent edge computing.

Key Responsibilities

• Develop and optimize embedded software and AI model inference pipelines for real-time object detection, tracking, and classification at the edge.
• Deploy and maintain vision-based deep learning models (YOLO, CNN, RNN, OCR) on resource-constrained devices using AI accelerators (e.g., Hailo, Jetson, Coral, ARM).
• Lead cloud-to-edge transformation of AI models, including quantization, pruning, conversion, and hardware-aware optimization.
• Collaborate with AI/ML researchers to integrate trained models into embedded applications and validate edge inference accuracy.
• Optimize for performance, latency, and power efficiency in real-world mobile or vehicle-mounted environments.
• Work with hardware abstraction layers, toolchains, and SDKs for AI acceleration (e.g., HailoRT, TensorRT, OpenVINO, ONNX, TFLite).
• Design and implement robust, fault-tolerant inference pipelines using tools like GStreamer, OpenCV, or FFmpeg.
• Contribute to system architecture discussions focused on scalable AI at the edge, real-time event capture, and safety-critical decision support.

Preferred Qualifications

• 5+ years of experience in embedded systems software and edge AI model deployment.
• Proficient in C/C++ and Python for embedded development and AI integration.
• Experience with quantization, pruning, and model conversion for edge inference (ONNX, TFLite, TensorRT, Dataflow Compiler etc.).
• Hands-on experience with at least one AI hardware platform (e.g., Hailo, NVIDIA Jetson, Intel Movidius, ARM Ethos-U, Google Coral).
• Solid understanding of deep learning architectures and model behavior (CNN, YOLOv5/v8, OCR, ReID, RNNs).
• Strong knowledge of video/image processing pipelines using GStreamer, OpenCV, or similar.
• Experience working with Linux-based embedded systems, build systems (Yocto, CMake), and cross-compilation toolchains.
• Understanding of IoT/vehicle edge environments, including system resource constraints, fault tolerance, and field deployment challenges.
• Bachelor’s or Master’s degree in Electrical Engineering, Computer Engineering, AI, or a related technical field.

Nice to Have

• Familiarity with multi-camera systems, object re-identification, or behavior/event detection at the edge.
• Knowledge of real-time messaging and telemetry systems (MQTT, ZeroMQ, or gRPC).
• Experience in public safety, ADAS, surveillance, or other vision-based safety-critical applications.
• Familiarity with data privacy, security, and compliance concerns for edge AI systems.

Why Join Us

• Be a core driver in bringing safety AI to the edge, where milliseconds matter.
• Help scale an intelligent, distributed AI platform that makes real-world impact across thousands of vehicles.
• Work at the intersection of AI, embedded systems, and public safety in a fast-paced, purpose-driven environment.
• Collaborate with engineers, data scientists, product managers, and field ops to build AI that directly saves lives.

Job Types: Full-time, Permanent

Experience:

• Embedded AI: 4 years (Preferred)
• Deep learning: 4 years (Preferred)
• Computer vision: 3 years (Preferred)
• Embedded software: 4 years (Preferred)

Work Location: In person