In short
Embedded software engineer deploying and optimizing on-device machine learning and computer-vision pipelines on resource-constrained edge hardware. Day-to-day involves integrating edge-AI into the Android camera/multimedia stack, bridging camera sensors with NPUs, and tuning memory, power, and latency. Core stack: C/C++, Python, TensorFlow Lite/ONNX Runtime, Android.
About the role
Key Notes:
- Strong experience in Embedded Software Engineering and C/C++ programming, with hands-on development of embedded systems.
- Practical experience deploying, integrating, and running Machine Learning (ML) models directly on edge devices.
- Experience in on-device ML performance optimization, improving model execution efficiency in resource-constrained embedded environments.
- Hands-on experience with power and latency optimization, reducing energy consumption and inference time during on-device ML execution.
- Strong understanding of embedded ML deployment, inference optimization, and hardware-software integration.
Responsibilities
- Design and deploy high-performance computer vision and machine learning pipelines directly onto resource-constrained edge hardware.
- Optimize concurrent ML workloads to maximize throughput and minimize latency when running multiple edge-AI features simultaneously.
- Integrate edge-AI algorithms seamlessly within the Android camera framework and multimedia subsystems.
- Collaborate across software and hardware teams to bridge low-level camera sensor subsystems with on-device neural processing units (NPUs).
- Profile and optimize memory footprint, power consumption, and execution timing to meet strict on-device performance constraints.
Minimum Qualifications
- Bachelor’s degree in computer science, Electrical Engineering, Computer Engineering, or a related technical field, or equivalent practical experience.
- Proficiency in Python, and C and C++ for low-level software development and optimization.
- Experience with machine learning frameworks optimized for edge deployment, such as TensorFlow Lite, ONNX Runtime, or PyTorch Edge.
- Familiarity with the Android platform stack, specifically regarding multimedia or camera subsystem integration.
- Solid understanding of computer vision fundamentals, digital signal processing (DSP), or basic camera architectures.
- Previous AI experience.
Preferred Qualifications
- Hands-on experience developing within the Android camera framework to control, process, route, and transform real-time image streams.
- Experience with model optimization techniques such as quantization, pruning, and hardware-specific compilation for NPUs, GPUs, or DSPs.
- Background in developing low-latency algorithms for real-time edge applications (e.g., biometric verification, image restoration, or on-device language processing).
- Familiarity with hardware-constrained vision pipelines, including specialized sensor integration and multi-threaded concurrency models.
- Hands-on experience debugging low-level software utilizing hardware profiling tools to isolate memory bottlenecks and scheduling conflicts.
- Former MAANG experience.