Staff Embedded ML/DSP Systems Engineer – Audio

About the role

Analog Devices is seeking a Staff Embedded ML/DSP Systems Engineer to lead the development of real‑time audio AI solutions. You will work at the intersection of hardware and software, shaping scalable, production‑grade AI/ML systems for hearable and wearable devices.

Key responsibilities

• Architect and optimize end‑to‑end deployment pipelines for compact audio AI models, covering quantization, profiling, and production deployment on DSP/NPU targets.
• Define DSP/NPU partitioning strategies, balancing workload, memory bandwidth, latency, and power across SoC processing elements.
• Develop bit‑exact reference models and collaborate with RTL teams for simulation‑to‑RTL validation and functional verification.
• Implement and optimize fixed‑point signal‑processing and neural‑network kernels for DSP and NPU instruction sets, maximizing MAC array and SIMD utilization.
• Profile and optimize inference performance under strict always‑on, real‑time constraints typical of hearable devices.
• Design and maintain model compression and quantization workflows (PTQ, QAT) with rigorous quality tracking.

Required profile

• Master’s or PhD in Electrical Engineering, signal processing, or a related field.
• 6+ years of audio/speech signal‑processing experience in a semiconductor environment with hands‑on DSP/NPU deployment.
• Proven expertise in fixed‑point algorithm implementation, model quantization, and cycle‑level optimization for resource‑constrained processors.
• Strong knowledge of simulation‑to‑RTL flows, bit‑exact modeling, and RTL co‑simulation.

Required skills

• C programming (embedded/firmware level)
• Python and MATLAB for algorithm development and profiling
• TensorFlow/TFLite, PyTorch/ONNX for model creation and conversion
• Fixed‑point DSP and NPU instruction‑set optimization
• Model quantization techniques (PTQ, QAT)
• Low‑level profiling tools and memory‑optimization strategies for embedded AI inference