Responsibilities
- Develop and maintain on-camera AI application pipelines, including object detection, event triggering, tracking, and camera peripheral control
- Port, convert, and optimize deep learning models for deployment across multiple SoC/NPU platforms while maintaining accuracy and real-time performance
- Integrate and validate AI inference engines with vendor-specific NPU SDKs and toolchains such as ONNX, TensorRT, QNN, or similar frameworks
- Collaborate with hardware and firmware teams on New Product Introduction (NPI) for next-generation AI cameras, from early SoC evaluation through production bring-up
- Train, fine-tune, and benchmark computer vision models for tasks such as detection, classification, action recognition, and video analytics
- Profile and debug performance bottlenecks across CPU, GPU, and NPU to meet latency, throughput, and power constraints
- Build and improve reusable frameworks, tools, and CI/CD pipelines for model conversion, testing, and deployment at scale
- Contribute to technical direction and system design for AI camera platforms; candidates with leadership experience may also help guide project execution and mentor engineers
Requirements
- MS or PhD in Computer Science, Electrical Engineering, or a related field or equivalent practical experience
- 3+ years of professional experience in edge AI system deployment
- Strong proficiency in Python and C/C++
- Hands-on experience with at least one deep learning framework such as PyTorch or TensorFlow
- Experience with model optimization techniques such as quantization, pruning, knowledge distillation, or compression for resource-constrained devices
- Strong English communication skills, with the ability to collaborate effectively across international and cross-functional teams
Nice to Have
- Experience in computer vision or NLP, including deep learning applications such as action recognition, anomaly detection, or video analytics
- Experience deploying deep learning models on embedded or edge platforms with NPU or AI accelerator hardware such as Ambarella, Qualcomm, InnoFusion, MediaTek, or similar SoCs
- Familiarity with model conversion toolchains and runtime engines such as TensorRT, ONNX Runtime, or SNPE
- Hands-on experience in NPI or product development lifecycle for camera or IoT hardware
- Familiarity with Large Language Models (LLMs) and Vision Language Models (VLMs) is a plus
- Experience with CI/CD, automated testing, or MLOps for edge AI deployment is a plus
- Demonstrated project ownership or mentoring experience is a plus
- Self-motivated, collaborative, and comfortable working in a fast-paced environment with challenging technical problems
