Edge AI Lead

Job Description

Role Summary: As the Edge AI Lead , you will oversee the design, optimization, and deployment of AI and machine learning models on edge devices, including NVIDIA Jetson, STM microcontrollers, and other embedded platforms. This role requires expertise in real-time operating systems (RTOS), the Robot Operating System (ROS), and hardware accelerators such as NPUs, TPUs, CPUs, and GPUs. You will lead a team to deliver high-performance, low-latency AI solutions and collaborate closely with robotics, embedded, and mechanical teams to create efficient, real-time autonomous systems. Core Responsibilities: 1. Design and Development of Edge AI Models Lead Edge AI Model Development: Drive the design and development of AI models optimized for edge devices and real-time processing. Focus on applications like object detection, SLAM (Simultaneous Localization and Mapping), sensor fusion, and predictive maintenance. Hardware-Aware Model Optimization: Develop models with an understanding of the constraints and capabilities of different hardware accelerators, including NPUs (Neural Processing Units), TPUs (Tensor Processing Units), CPUs, and GPUs. Optimize models through quantization, pruning, and knowledge distillation to fit the processing capabilities of edge devices. Integration with RTOS and ROS: Ensure models are compatible with real-time operating systems (RTOS) for reliable, time-sensitive applications. Leverage ROS for seamless integration with robotic systems, enabling efficient communication between AI models and control systems. 2. Deployment and Integration of Edge AI Models Edge Device Deployment: Lead the deployment of AI models on edge devices, including platforms like NVIDIA Jetson, STM32 microcontrollers, ARM Cortex, and Qualcomm Snapdragon. Ensure seamless integration with RTOS and sensor interfaces for reliable real-time processing. Coordinate with Robotics and Control Teams: Collaborate with robotics and embedded control teams to ensure AI models are effectively integrated within robotic platforms. Address challenges related to sensor placement, communication protocols, and power distribution. Optimize Data Flow with ROS and RTOS: Ensure AI models are well-integrated with ROS for real-time data exchange and with RTOS for deterministic performance. Optimize communication pathways between sensors, processors, and AI models for smooth data flow. 3. Edge AI Pipeline and MLOps Implement MLOps for Edge AI: Develop MLOps pipelines specifically tailored for edge deployment, including CI/CD, automated testing, and performance monitoring. Ensure models can be continuously updated, retrained, and redeployed based on operational data. Model Lifecycle Management: Manage model versioning, deployment, and rollback strategies to ensure reliable, up-to-date models are deployed on edge devices. Develop feedback loops to monitor model performance and enable continuous improvement. Real-Time Monitoring and Troubleshooting: Implement monitoring tools to track metrics such as latency, accuracy, and resource usage in real time. Develop protocols for troubleshooting issues to maintain reliability and performance in production environments. 4. Hardware Acceleration and Performance Optimization Leverage Hardware Accelerators: Use tools like NVIDIA TensorRT, CUDA, OpenVINO, and Qualcomm AI Stack to maximize performance on GPUs, TPUs, NPUs, and other hardware accelerators. Optimize AI inference to take full advantage of the specific architecture of each edge device. Low-Level Programming for Efficiency: Utilize low-level programming languages like C, C++, and assembly to optimize code execution on embedded systems. Implement custom kernels and optimize memory usage to improve performance on constrained hardware. Energy and Power Optimization: Focus on energy-efficient model deployment to maximize battery life and reduce power consumption, particularly for mobile and autonomous robots. Apply power management strategies to balance performance and energy efficiency. 5. Leadership and Team Management Lead and Mentor the Edge AI Engineering Team: Build and guide a team of Edge AI engineers, providing technical direction, mentorship, and hands-on support. Set clear performance expectations, provide regular feedback, and support professional growth. Project and Resource Management: Oversee project timelines, resource allocation, and task prioritization for the Edge AI team. Ensure alignment with broader organizational goals and coordinate with cross-functional teams to achieve project milestones. Foster a Culture of Innovation and Technical Excellence: Promote a culture of continuous learning, experimentation, and technical rigor within the team. Encourage knowledge-sharing and collaboration to solve complex edge AI challenges. 6. Collaboration with Cross-Functional Teams Work Closely with Embedded Systems and Software Teams: Collaborate with embedded systems and software engineers to ensure compatibility of AI models with embedded platforms, RTOS, and other system software. Address integration challenges related to memory constraints and real-time requirements. Coordinate with Robotics, Data, and Mechanical Teams: Work closely with robotics engineers, data scientists, and mechanical engineers to align on data needs, model integration, and design constraints. Ensure that Edge AI solutions support and enhance the capabilities of the robotic systems. Align with Head Robotics and AI Software Leadership: Collaborate with the Head Robotics and AGI/DL Software Head to ensure Edge AI development aligns with broader AI and robotics objectives. Ensure edge solutions support real-time perception, decision-making, and autonomous functionality. 7. Innovation and Research in Edge AI and Embedded Systems Explore Emerging Edge AI Technologies: Stay current with advancements in edge AI hardware, software, and frameworks, including low-power inference engines, neural accelerators, and lightweight edge computing frameworks. Assess and implement new technologies to enhance performance and efficiency. Lead R&D Projects on Edge AI Challenges: Initiate research projects to address unique edge AI challenges such as low-power image recognition, real-time SLAM, and multi-sensor fusion on edge devices. Lead proof-of-concept projects to validate new approaches and methodologies. Contribute to Open-Source and Industry Engagement: Encourage team participation in open-source projects and engagement with the AI/ML community. Promote contributions to Edge AI tools, libraries, and frameworks to stay connected with industry trends and drive community-driven innovation. 8. Compliance, Safety, and Quality Assurance Ensure Safety and Compliance Standards: Work closely with quality and compliance teams to ensure that Edge AI models meet industry safety standards, particularly for autonomous and industrial applications. Quality Control and Documentation: Implement quality control processes for Edge AI models, ensuring reliability, consistency, and long-term performance. Maintain comprehensive documentation on deployment protocols, optimization techniques, and troubleshooting procedures. Develop and Enforce Best Practices: Create and maintain best practices for edge AI development, deployment, and maintenance. Ensure protocols are well-documented, accessible, and regularly updated. Required Qualifications: Education: Bachelor's or Master's degree in Computer Science, Electrical Engineering, Robotics, or a related field. Advanced degrees or certifications in AI, embedded systems, or robotics are preferred. Experience: 8+ years of experience in AI, machine learning, or embedded systems, with at least 3 years in edge computing or embedded AI. Demonstrated experience deploying and optimizing AI/ML models on edge devices, with a focus on low-latency, high-performance environments. Strong experience in leading teams and managing projects involving edge AI, embedded systems, or robotics. Technical Skills: Machine Learning and AI: Proficiency in ML/DL frameworks (e.g., TensorFlow Lite, PyTorch, ONNX) with extensive experience in model optimization for edge devices. Real-Time Systems (RTOS) and ROS: Experience working with real-time operating systems (RTOS) and the Robot Operating System (ROS) to enable reliable real-time processing and seamless communication in robotic systems. Hardware Acceleration and Low-Level Programming: Expertise in using tools like TensorRT, CUDA, OpenVINO, and low-level programming languages (e.g., C, C++) to optimize AI inference on hardware accelerators (NPU, TPU, CPU, GPU). Embedded Systems and Edge Platforms: Deep knowledge of edge computing platforms such as NVIDIA Jetson, STM32 microcontrollers, ARM Cortex, and Qualcomm Snapdragon. Understanding of RTOS, middleware, and embedded programming. Preferred Qualifications: Robotics and Autonomous Systems Knowledge: Understanding of robotics control systems, perception systems, and integration of AI within autonomous systems. Energy-Efficient AI: Experience with optimizing AI models for low-power applications, particularly for mobile and autonomous robotics where power consumption is critical. Project Management and Agile Practices: Familiarity with Agile methodologies and experience in project management to oversee complex, cross-functional projects effectively. Open-Source and Community Engagement: A track record of contributions to open-source AI/ML or embedded systems projects and active participation in the Edge AI community.